Skip to main content

Full text of "the Text of the New Testament 2nd Edit"

See other formats



Its Transmission, Corruption, and 



Professor of Mew Testament Language and Literature 
Princeton Theological Seminary 



Preface to the Second Edition 

During the four years that have elapsed since the initial 
publication of this book in 1964, a great amount of 
textual research has continued to come from the presses 
in both Europe and America. References to some of these 
publications were included in the German translation of the 
volume issued in 1966 under the title Der Text des Neuen Test- 
aments; Einftihrung in die neutestamentliche Textkritik (Kohlhammer 
Verlag, Stuttgart) . The second printing of the English edition 
provides opportunity to introduce a variety of small alterations 
throughout the volume as well as to include references to more 
than one hundred and fifty books and articles dealing with 
Greek manuscripts, early versions, and textual studies of recently 
discovered witnesses to the text of the New Testament. In 
order not to disturb the pagination, most of the new material 
has been placed at the close of the book (pp. 261-73), to which 
the reader's attention is directed by appropriate cross references. 

February ig68 


The necessity of applying textual criticism to the books of 
the New Testament arises from two circumstances : (a) none 
of the original documents is extant, and (b) the existing 
copies differ from one another. The textual critic seeks to ascer- 
tain from the divergent copies which form of the text should 
be regarded as most nearly conforming to the original. In some 
cases the evidence will be found to be so evenly divided that it 
is extremely difficult to decide between two variant readings. 
In other instances, however, the critic can arrive at a decision 
based on more or less compelling reasons for preferring one 
reading and rejecting another. 

It is the purpose of this book to supply the student with 
information concerning both the science and the art of textual 


criticism as applied to the New Testament. The science of 
textual criticism deals with (a) the making and transmission of 
ancient manuscripts, (b) the description of the most important 
witnesses to the New Testament text, and (c) the history of the 
textual criticism of the New Testament as reflected in the succes- 
sion of printed editions of the Greek Testament. The art of 
textual criticism refers to the application of reasoned considera- 
tions in choosing among variant readings. The results of the 
practice of textual criticism have differed from one generation to 
another, partly because the balance in the quantity and the 
quality of witnesses available has gradually altered owing to 
the acquisition of additional manuscripts, and partly because 
theories and procedures of evaluating textual evidence have 
varied over the years. In Part Three of the volume the author has 
attempted to provide a succinct account of the several schools 
of textual methodology, indicating at the same time what in his 
view is the most satisfactory critical procedure for beginners 
to follow. 

The author gratefully acknowledges the courtesy of the 
following for granting permission to reproduce specimen folios 
and diagrams of manuscripts : Bibliotheque Bodmer, the Bod- 
leian Library, the British Museum, the Cambridge University 
Press, Dr. W. H. P. Hatch, Macmillan & Co., Ltd., and the 
Speer Library of Princeton Theological Seminary. 

Finally, I should like to express my gratitude to the Delegates 
of the Oxford University Press for their acceptance of my book 
for publication. I am also indebted to the readers of the Press 
for their customary care and painstaking vigilance in the read- 
ing of the proofs. 


Princeton, New Jersey 
August 1963 

List of Figures 

i. Greek Uncial Script (codex Sinai ticus) page 10 

2. Greek Minuscule Script (lectionary 303) 1 1 

3. Stemma Illustrating Streeter's Theory of Local Texts 171 

4. Chart of Witnesses and the Local Texts 172 

List of Plates 

(at end) 

i. a. Limestone statuette of Egyptian scribe 
b. Codex Amiatinus, Ezra the scribe 

ii. Chester Beatty Biblical Papyrus II (p 46 ) 
in. Bodmer Papyrus XIV (p? 5 ) 
iv. Codex Sinaiticus (X) 

v. Codex Bezae (D) 

vi. a. Codex Laudianus (E») 

b. Parchment leaf of Book of Revelation (0169) 

vii. Codex Rossanensis (2), Christ and Barabbas before Pilate 

viii. Codex Basiliensis (E) 

ix. Codex Mosquensis (K a P) 

x. Greek Gospel Lectionary 56a 

xi. Greek Gospel MS. 274, with intermediate ending of Mark 

xii. a. Curetonian Syriac MS. 
b. Gothic codex Argenteus 

xiii. a. Codex Sangallensis (A) 
b. Codex Gigas 

xiv. a. Chester Beatty Coptic (Sahidic) MS. B 
b. Armenian Gospel MS. 229, Etchmiadzin 

xv. Greek Gospel MS. 2 

xvi. Complutensian Polyglot Bible 

The Making of Ancient Books 

Until the invention of printing with movable type in the 
fifteenth century the text of the New Testament — and, 
indeed, the text of eyery ancient record — could be trans- 
mitted only by laboriously copying it letter by letter and word 
by word. The consideration, therefore, of the processes involved 
in the making and transcribing of manuscripts is of the utmost 
importance to the historian of ancient culture in general and to 
the student of the New Testament in particular. The following 
sections deal with those aspects of Greek palaeography 1 that 
bear upon the textual criticism of the New Testament. 


Clay tablets, stone, bone, wood, leather, various metals, pot- 
sherds (ostraca) , papyrus, and parchment (vellum) were all used 
in antiquity to receive writing. Among these several materials, 
the student of the New Testament is interested chiefly in the 
last two, for almost all New Testament manuscripts are made of 
either papyrus or parchment. 

The manufacture of papyrus was a flourishing business in 
Egypt, for the papyrus plant grew plentifully in the shallow 
waters of the Nile at the delta (cf. Job viii. n, 'Can papyrus 
grow where there is no marsh?'). About 12 or 15 feet in height, 
the stem of the plant, which was triangular in cross-section and 
as thick as a man's wrist, was cut into sections about a foot long. 
Each section was split open lengthwise and the pith cut into 
thin strips. A layer of these was placed on a flat surface, all the 

1 Standard works on Greek palaeography include, for example, Viktor Gardt- 
hausen, Griechische Palaeographie, 2 vols., 2te Aufl. (Leipzig, 191 1-13) ; E. M. Thomp- 
son, An Intrdduction to Greek and Latin Palaeography (Oxford, 1912); A. Sigalas, 
'Iaropia tt/j 'EXXriviieijs rpaifjfjs (Thessaloniki, 1934); L. Gonzaga da Fonseca, 
S.J., Epitome introductions in palaeographiam Graecam (Biblicam), ed. altera (Rome, 
1944) ; and B. A. van Groningen, Short Manual of Greek Palaeography (Leiden, 1940; 
3rd ed., 1963). For additional bibliography, see the present writer's article 'Palaeo- 
graphy' in Encyclopedia Americana, xxi (1958), pp. 163-6. See below, p. 261. 

4 The Making of Ancient Books 

fibres running in the same direction, and on top another layer 
was laid, with the fibres running at right angles to the lower 
layer. The two layers were then pressed together until they 
formed one fabric — a fabric which, though now so brittle that it 
can sometimes be crumbled into powder, once had a strength 
nearly equal to that of good paper. 

The manufacture of parchment for writing purposes has an 
interesting history. According to Pliny the Elder (in his Natural 
History, xiii. 21 f.), it was King Eumenes of Pergamum, a city 
in Mysia of Asia Minor, who promoted the preparation and use 
of parchment. This ruler (probably Eumenes II, who ruled 
from 197 to 159 B.C.) planned to found a library in his city 
which would rival the famous library of Alexandria. This 
ambition did not please Ptolemy of Egypt (probably Ptolemy 
Epiphanes, 205-182 B.C.), who thereupon put an embargo on 
the export of papyrus sections. It was this embargo which 
forced Eumenes to develop the production of vellum, which 
from the place of its origin received the Greek name -rrepyafirjvri 
(whence our English word 'parchment' is derived). Whatever 
may be thought of the details of this story — actually leather 
(parchment) was used for books long before Eumenes — the 
gist of it is probably true, namely that a high quality of parch- 
ment was developed at Pergamum, and that the city became 
famous in the manufacture and export of this kind of writing 
material, eventually giving its name to the product. 

Parchment or vellum (the two words are often used inter- 
changeably, but exact writers restrict the word 'vellum' to de- 
scribe a finer, superior quality of parchment) was made from 
the skins of cattle, sheep, goats, and antelopes, and especially 
from the young of these animals. After the hair had been re- 
moved by scraping, the skins were washed, smoothed with 
pumice, and dressed with chalk. De luxe editions, according to 
St. Jerome, who did not approve of such extravagance, 1 were 

■ In his famous letter to Eustochium, Jerome inveighs against anomalous extra- 
vagance: 'Parchments are dyed purple, gold is melted into lettering, manuscripts 
are decked with jewels, while Christ lies at the door naked and dying' (Epist. xxii. 
32 ; cf. also Jerome's preface to the Book of Job, and see Evaristo Arns, La Technique 
du livre d'apris Saint Jerome [Paris, 1953]). Writing to a correspondent named Laeta, 
who has asked how she ought to rear her young daughter, he advises, 'Let her 
treasures be not gems or silks, but manuscripts of the holy Scriptures ; and in these 
let her think less of gilding and Babylonian parchment and arabesque patterns, 
than of correctness and accurate punctuation' {Epist. cvii. ia). For a list of extant 

The Making of Ancient Books 5 

made of vellum dyed a deep purple and written with gold and 
silver inks. Ordinary editions were written with black or brown 
ink and had decorative headings and initial letters coloured 
with blue or yellow or (most often) red ink — whence the word 
'rubric', from ruber, the Latin for 'red'. 

Vellum or parchment continued to be generally used until 
the late Middle Ages. At that time paper, which was made of 
cotton, hemp, or flax, having been introduced into Europe 
from China by Arabian traders, became popular and sup- 
planted other writing materials. (See below, p. 261.) 


In the Graeco-Roman world literary works were customarily 
published in the format of a scroll, made of papyrus or parch- 
ment. The papyrus roll or scroll was made by gluing together, 
side by side, separate sheets of papyrus and then winding the 
long strip around a stick, thus producing a volume (a word 
derived from the Latin volumen, 'something rolled up'). The 
length of such a papyrus roll was limited by considerations of 
convenience in handling the roll ; the normal Greek literary roll 
seldom exceeded 35 feet in length. 2 Ancient authors therefore 
would divide a long literary work into several 'books', each of 
which could be accommodated in one roll. The two longest 
books in the New Testament — the Gospel of Luke and the Book 
of Acts — would each have filled an ordinary papyrus roll of 

purple manuscripts of the Greek and Latin Bible, as well as a discussion of the 
special skills required to produce a codex aureus purpureas, see E. A. Lowe in Studies in 
Art and Literature for Belle da Costa Greene, ed. by Dorothy Miner (Princeton, 1954), 
pp. 266-8. 

1 See F. G. Kenyon, Books and Readers in Ancient Greece and Rome (Oxford, 1932; 
2nd ed., 1951) ; Henry A. Sanders, 'The Beginnings of the Modern Book', Michigan 
Alumnus Review, xliv (1938), pp. 95-1 1 1 ; C. C. McCown, 'Codex and Roll in the 
New Testament', Harvard Theological Review, xxxiv (1941), pp. 219-50; id., 'The 
Earliest Christian Books', Biblical Archaeologist, vi (1943), pp. 21-31; Frank W. 
Beare, 'Books and Publication in the Ancient World', University of Toronto Quarterly, 
xiv (1945), pp. 150-67; H. L. Pinner, The World of Books in Classical Antiquity 
(Leiden, 1948); C. H. Roberts, 'The Codex', Proceedings of the British Academy, xl 
(1954), pp. 169-204; and Robert Devreesse, Introduction a I'etude des manuscrits grecs 
(Paris, 1954). See below, p. 261. 

2 Calli machus, the learned cataloguer of books in the great library at Alexandria, 
was accustomed to say, 'A big book is a big nuisance' (ii-tya /3i/3Ai'ox /J-tya kclkov). 
Ceremonial copies of the Egyptian Book of the Dead have been found which 
measure more than 100 feet in length, but these were not meant to be read but to 
be buried in the tomb of a rich owner. 

82S158 B 

6 The Making of Ancient Books 

31 or 32 feet in length. Doubtless this is one of the reasons why 
Luke- Acts was issued in two volumes instead of one. 

On the roll thus formed the writing was arranged in a series 
of columns, each about 2 or 3 inches wide. The height of the 
columns, which ran parallel to the stick on which the roll was 
wound, varied, of course, with the height of the original papyrus 
sheets. Sometimes, but not very often, the roll was written on 
both sides (see Rev. v. 1) ; this was called an opisthograph. 

The roll was relatively inconvenient to use. The reader had 
to employ both hands, unrolling the scroll with one hand and 
rolling it up with the other as the reading proceeded. Moreover, 
the Christian community soon discovered how laborious it was 
to try to find specific passages in their sacred books when they 
were written in roll-form. Early in the second century (or per- 
haps even at the close of the first century) the codex, or leaf- 
form of book, began to come into extensive use in the Church. 
A codex was made by folding one or more sheets of papyrus in 
the middle and sewing them together. Christians found that this 
form had a number of advantages over the roll: (a) it permitted 
all four Gospels or all the Epistles of Paul to be bound into one 
book, a format which was impossible so long as the roll was 
used; (b) it facilitated the consultation of proof-texts ; (c) it was 
better adapted to receiving writing on both sides of the page, 
thus keeping the cost of production down. The suggestion may 
well be true 1 that it was Gentile Christians who early adopted 
the codex-form for their Scriptures, instead of the roll-form, as 
part of a deliberate attempt to differentiate the usage of the 
Church from that of the synagogue, which was accustomed to 
transmit the Old Testament on scrolls. 

The advantages of parchment over papyrus for the making 
of books seem obvious to us today. It was a much tougher and 
more lasting material than the more fragile papyrus. Moreover, 
parchment leaves could receive writing without difficulty on 
both sides, whereas the vertical direction of the fibres on the 
'reverse', side of a sheet of papyrus made that side somewhat less 
satisfactory than the other as a writing surface. On the other 
hand, parchment also had its defects. For example, the edges 
of parchment leaves are apt to become puckered and uneven. 

1 See Peter Katz, 'The Early Christians' Use of Codices instead of Rolls', 
Journal of Theological Studies, xliv (1945), pp. 63-65. See below, p. 261. 

The Making of Ancient Books 7 

Furthermore, according to the observation of Galen, the famous 
Greek physician of the second century a.d., parchment, which 
is shiny, strains the eyes more than does papyrus, which does not 
reflect so much light. 

Eusebius, the noted Christian scholar of Caesarea in Palestine, 
included in his Life of Constantine information concerning an im- 
perial request for fifty parchment manuscripts. About a.d. 331, 
when Constantine wished to secure copies of the Scriptures for 
the new churches which he proposed to build in Constantinople, 
he wrote to Eusebius requesting him to arrange without delay 
for the production of 'fifty copies of the sacred Scriptures . . . 
to be written on fine parchment in a legible manner, and in a 
convenient portable form, by professional scribes {KaX\vypa<f>oi) 
thoroughly accomplished in their art'. 1 These orders, Eusebius 
continues, 'were followed by the immediate execution of the 
work itself, which we sent him in magnificent and elaborately 
bound volumes of threefold and fourfold forms'. 2 

The suggestion has been made by several scholars that the 
two oldest parchment manuscripts of the Bible which are in 
existence today, namely codex Vaticanus and codex Sinaiticus 
(see Chapter II for descriptions of these manuscripts), may have 
been among those ordered by Constantine. It has been pointed 
out that Eusebius' curious expression, 'volumes of threefold 
and fourfold forms', agrees with the circumstance that these two 
codices have respectively three columns and four columns on 
each page. There are, however, one or two indications which 

1 Eusebius, Life of Constantine, iv. 36. 

1 The Greek text of the concluding clause (ex mXvreXws ^aitrjfievois Ttixco-w 
rpiaoa kcll rerpaaaa hiaiT€[iifidvTujv rj^idiv) is difficult to interpret, and the words 
Tpiaaa koX rerpaaad have been taken in widely different senses. Thus it has been 
suggested that the words refer to codices which were composed of quires of three or 
four double leaves; that they were polyglot Bibles in three or four languages; that 
they were harmonies of three or four Gospels; that copies were sent off to Constan- 
tine three or four at a time; that each Bible was in three or four parts; or that the 
pages had three or four columns of script. Each of these interpretations involves 
more or less serious difficulties ; perhaps the least unsatisfactory interpretation is 
the one mentioned last. For discussions of the problems involved, see Kirsopp Lake, 
'The Sinaitic and Vatican Manuscripts and the Copies sent by Eusebius to 
Constantinople', Harvard Theological Review, xi (1918), pp. 32-35; J. H. Ropes, 
The Text of Acts (= The Beginnings of Christianity, part I, vol. iii; London, 1926), 
pp. xxxvi ff ; Carl Wendel, 'Der Bibel-Auftrag Kaiser Konstantins', ^entralblatt 
fur Bibliothekswesen, lvi (1939), pp. 165-75; and T. C. Skeat, 'The Use of Dicta- 
tion in Ancient Book-Production', Proceedings of the British Academy, xlii (1956), 
pp. 196 f- 

8 The Making of Ancient Books 

point to Egypt as the place of origin of codex Vaticanus, and the 
type of text found in both codices is unlike that used by Eusebius. 
The most that can be said with certainty, therefore, is that codices 
Vaticanus and Sinaiticus are doubtless like those which Constan- 
tine ordered Eusebius to have copied. 


In writing on papyrus the scribe was accustomed to utilize 
the' horizontal fibres on the recto side of the sheet as guide lines 
for his script. Before writing on parchment he would score the 
surface with a blunt-pointed instrument, drawing not only 
horizontal lines but two or more vertical lines as well, thus 
marking the margins of each column of writing. In many manu- 
scripts these guide lines are still visible, as are also the pinpricks 
which the scribe made first in order to guide him in ruling 
the vellum. 1 Different schools of scribes employed various pro- 
cedures of ruling, and occasionally it is possible for the modern 
scholar to identify the place of origin of a newly discovered 
manuscript by comparing its ruling pattern (as it is called) with 
those in manuscripts whose place of origin is known. 2 Since the 
hair side of vellum is darker than the flesh side, it was found that 
the most pleasing effect upon the reader was obtained only when 
the separate sheets were not indiscriminately gathered together 
in quires, but when the hair side of one page faced the hair side 
of the opposite page, and the flesh side faced the flesh side, 
wherever the book was opened. 3 

In antiquity two styles of Greek handwriting were in general 

1 There is even a science of pinpricks! See E. K. Rand, 'Prickings in a Manu- 
script of Orleans', Transactions and Proceedings of the American Philological Association, 
lxx (1939), pp. 327-41 ; and L. W.Jones, ' "Pin Pricks" at the Morgan Library', 
ibid., pp. 318-26; 'Where are the Prickings?', op. cit., lxxii (1944), pp. 71-86; 
'Pricking Manuscripts: the Instruments and their Significance', Speculum, a Journal 
of Mediaeval Studies, xxi (1946), pp. 389-403; and 'Prickings as Clues to Date and 
Origin: the Eighth Century', Medievalia el humanistica, xiv (1962), pp. 15-22. Rand 
dealt earlier with various methods of ruling manuscripts in vogue during the Middle 
Ages; see his study entitled 'How Many Leaves at a Time?' in Palaeographica Lalina, 
v (1927). PP- 52-78. 

1 For a list of several hundred different ruling patterns, see Kirsopp and Silva 
Lake, Dated Greek Minuscule Manuscripts to the Tear 1200 A.D. (Boston, 1934-45). See 
below, p. 261. 

3 This characteristic feature of parchment codices was discovered towards the 
end of the last century by Caspar R. Gregory; see his article 'The Quires in Greek 
Manuscripts', American Journal of Philology, vii (1886), pp. 27-32. 

The Making of Ancient Books 9 

use. The cursive or 'running' hand, which could be written 
rapidly, was employed for non-literary, everyday documents, 
such as letters, accounts, receipts, petitions, deeds, and the like. 
Contractions and abbreviations of frequently recurring words 
(such as the definite article and certain prepositions) were 
common. Literary works, on the other hand, were written in a 
more formal style of handwriting, called uncials. 1 This 'book- 
hand' was characterized by more deliberate and carefully 
executed letters, each one separate from the others, somewhat 
like our capital letters. Some of the most beautiful specimens of 
Greek handwriting are certain classical and Biblical manuscripts 
dating from the third to the sixth century. In the course of time, 
however, the style of the book-hand began to deteriorate, and 
uncials became thick and clumsy. Then, about the beginning of 
the ninth century, a reform in handwriting was initiated, and a 
script of smaller letters in a running hand, called minuscules, 2 
was created for the production of books. 3 This modified form of 
the cursive script became popular almost at once throughout the 
Greek world, though some liturgical books continued for one or 
two centuries to be written in uncial script. Thus manuscripts 

1 The word 'uncial' is derived from the Latin uncia, meaning 'a twelfth part' 
of anything. Apparently the term came to be applied to letters which occupied 
roughly about one-twelfth of an ordinary line of writing. Cf. W. H. P. Hatch, 
'The Origin and Meaning of the Term "Uncial" ', Classical Philology, xxx (1935), 

PP- 247-54- 

2 The word 'minuscule' is derived from the Latin minusculus, meaning 'rather 
small'; it was used because the letters designated by this term are usually smaller 
than capitals and uncials. 

3 The credit for initiating the reform in Greek handwriting has been commonly 
attributed to scholarly monks at the monastery of the Studium in Constantinople 
(see, for example, G. Zereteli, 'Wo ist das Tetraevangelium von Porphyrius 
Uspenskij aus dem Jahre 835 entstanden?', Byzantinische ^eitschrift, ix [igoo], 
pp. 649-53, and T. W. Allen, 'The Origin of the Greek Minuscule Hand', Journal 
of Hellenic Studies, xl [1920], pp. 1-12), but recently it has been argued that the 
perfecting of minuscule script for book production was the work of humanistic 
scholars who were involved in the revival of culture at Constantinople during the 
second epoch of iconoclasm (see Bertrand Hemmerdinger, Essai sur I'histoire du texte 
de Thucydide [Paris, 1955], pp. 33"39)- 

It may be mentioned that the earliest known minuscule Greek manuscript bearing 
a date is a copy of the four Gospels, now in the Public Library of Leningrad, with 
its all-important colophon dated 7 May 6343 (= a.d. 835) by the monk Nicolaus, 
later abbot of the Studium. This manuscript has raised problems for scholars; 
palaeographically the handwriting appears to be too mature and fully developed to 
stand at the beginning of the minuscule period, yet no forerunners have been recog- 
nized among extant manuscripts. For a discussion, see Aubrey Diller, 'A Com- 
panion to the Uspenski Gospels', Byzantinische ^eitschrift, xlix (1956), pp. 332-5. 


The Making of Ancient Books 

fall into two rather well-defined groups, the earlier being writ- 
ten in uncial letters (see Fig. i) and the later in minuscules (see 
Fig. 2). 

n iTxneTftDj^Mo 
noyoyK^i xgnth 
cuceaxNeri x<fnxi 



M H € X CI Nf I ZKKli 


AxxxAee r 1 eceNf 


XKXNexi KAierrHi 

d>HTi x h cxioyn 
XKoycereK a i-y 


r xen o ki recfcx-.. 


rxf h k xrx 1 xroy 


rxfecocH Koys 
kx i*Tr>y CO<f>eAX 
M OyCAYTXl J N €**" 
^^ *W£ ' c KAIFfc 

d>ci M>Koycuxr 




Kyroyc ,-**„ 

Fig. 1. Greek Uncial Script (from codex Sinaiticus, 4th cent.) 
(actual width of each column about 2j in.) 

Col. a, Matt. xiii. 5-10, aAAa 8e €tt€V€i> €\ttl ra 77-^00)877 o\ttov ovk £ix* v yv v I 7r °M 7 ? 1 ' 
Kai €v$€\ios e^ai'eTtAei' 8c|a to ^7/ t^c? fiaOos \ yt)S 7/Aiou oc apaJTiAaPTo? €Kavfia\Tta6r} 
Kai 81a to I [IT] €X €LV pi£ae t|£T>pai'6 ) 7/ | aAAa 8e f7rfa«' t Irn tw aKavdas \ Kai avefirjaav 
at I aKavOai Kai c-nvi\£av aura | aAAa 8e €7Tfafi/ f|7Ti t7/p yTii' t7/i> Ka^T/p *ai tSiSou j 
Kaprrov o ^iev e |KaToi> o 8e ff7iKoi'|Ta o Se A o exa)? [ a>Ta [insert aKoveiv from the left- 
hand margin] aKovfroj \ Kai rrpoo(A8ovT(S || col. b, Matt. xiii. 14-16 Kai avairX-qpov] 
rai avrois f) Trpo|^7iTia -qoa'iov f) \ Xcyovoa aKo-q \ clkovo€T€ Kai ov \ p.-q avv-qre Kai | j3Ae- 
rrovr€S /9Af|t/r7jTe Kai ou p.rj iJSt/tc €rraxvv6ij \ yap rj Kaphia rov \ Xaov toutov Kai | tois 
woiv auTiov | fiaptios i)Kovoav \ Kai rovs o(f>6aX\^.ovs olvtcov €Ka^i\yLVaav y.iynoTt \ 
'ihoioiv toi? o\(f>da\[iois Kai tois | cooiv [avriov between the lines] aKouaajaiv \ Kai ttj 
Kaphia avvio\aiv Kai emoTpel'piooiv Kai iaaop.e \ avTovs \ vpiov 8e paKapi. The Eusebian 
canon numerals (see pp. 24 f. below) stand between the columns /pA8 _ 1 34\ 

I « ' 5 /" 
The advantages of using minuscule script are obvious. Minus- 
cule letters, as their name suggests, are smaller than uncials, 
and thus the writing is more compact. Hence, when the minus- 


xAXXAfeen eceNe 




icxroNpxe^zH * 


Kx»rrroc^xe-Nf. *T 

The Making of Ancient Books 1 1 

cule hand was used, less parchment was required and therefore 
the book was more economical. Furthermore, a literary work 
could be produced which was less bulky and therefore easier 
to handle than a larger manuscript. Moreover, it was possible 
to write minuscule letters more rapidly than uncials, and con- 
sequently books could be produced more quickly and more 

l^m ojl/top axp ay tt> or T ^JSt a»luxipoDo-j<»iya»' 
cryoy ero oltco-ouutco y • 'Iti an*l*-iiof 0004/ p war 
Ia*u tf-oDTjynrpoaa^ I t<ou er>c*TtDyfu*3'i/V 

o\ajMytxDy\juauio\t-&- • crpjJ4_$-rour«TTDjv» *"»>** 

• •* V * " ** i " ' 

Jl « lAJJJ 0>|/ * |A4 JJ . CD OT f p»«-ODXUTT>VlU«TI * ^<^T»" 

\ <v\ c i IajLu if cnj-rvi o 1 d\ ooLijLif oo«TPV\9v • 

ft« >.i' _ » \ , • * • » , „ 

d\x> \jutxcoa o\M\to\y6 UxuNluiixrdEfsaxrrDV 

k*Ym V tout vp clA cutr ; o i a\jo yuxSn-TrxiX cu 

U-OU Ct^l OU3P0LJ/ •t-db' OULJ ^01Xi**TDey.l*X14UllijO 

•TTt «-m Co p ou vrcLf arp o I j> inr* o-cLy * iu > \\j " 

djp otuju h \9p o i cr fi* ' aLcna4*-&-jj o or olutd vor 

JJ O UO" TO \XT &y <»Wjlaj odxO^O VrQ-OtlXf -r rfn - 

Fig. 2. Greek Minuscule Script (from lectionary 303, 12th or 13th cent.) 
(actual width of each column about 3 J in.) 

Col. a, Luke xxiv. 31—33, Kai avros atfravros | eyevero air auTajv | Kai etirov irpos 
aAjA^Aou?' ovxl 1) /cap[8i'a rj/xcuv Kaio/j.e\v7) t}v cV T)ixiv. cos ejAaAet ^jlaip eV ttj d|8ai /cai aij 
Snjpoiyep ] ^if tcls ypa<f>ds ; ] Kai apaffTapTes au|T^ t^ wpa vneaTpe\ipav els i[epouffa]- 
AtJ^a* icai I edpov <jvvr}6poirOp.4\vovs tovs cf8e/ca || col. b, John i. 35—38, Tw Kaipaj 
cKeivw I eia-nj/cei 6 Lwdvvys I /cat e/c tojp juaaiyjTajp auTOu Suo- icai [ epfiXeipas tw 
i[7jao]u 7re|pifl-aToufTi* Aeycf | iStf d a^vo? tov 8[€o]v- \ Kai tJkovoclv avrov \ ol 8uo 
padrjTai \a\\ovvTos. /cat r)Ko\\ovdT]oav toj i[iyao]u' | OTpatfreis 6 ifiyaou]? *ai 6c\aad[jL£vos 
auTous I d*oAou0oOpTas. 

cheaply. It is easy to understand that this change in the style of 
script had a profound effect upon the textual tradition of the 
Greek Bible. Now the possession of copies of the Scriptures (and 
of other literary works) was placed within reach of persons of 
limited means. When literary works were copied almost exclu- 
sively in the uncial script, such persons were obliged to get along 

12 The Making of Ancient Books 

without many books. Thus the minuscule hand was an impor- 
tant factor in the dissemination of culture in general and of the 
Scriptures in particular. The minuscule manuscripts of the New 
Testament outnumber the uncial manuscripts by more than ten 
to one, and although one must make allowance for the greater 
antiquity of the uncial style (and consequently the greater 
likelihood of the destruction of uncial manuscripts through the 
ravages of time) , very much of the disparity in the number of 
the survivors must be due to the increased ease with which the 
minuscule copies could be produced. 

In times of economic depression, when the cost of vellum 
increased, the parchment of an older manuscript would be used 
over again. The original writing was scraped and washed off, 
the surface re-smoothed, and the new literary material written 
on the salvaged material. Such a book was called a palimpsest 
(which means 're-scraped', from irdXiv and ipdoj). One of the 
half-dozen or so most important parchment manuscripts of the 
New Testament is such a palimpsest ; its name is codex Ephraemi 
rescriptus. Written in the fifth century, it was erased in the 
twelfth century and many of the sheets rewritten with the text 
of a Greek translation of thirty-eight treatises or sermons by 
St. Ephraem, a Syrian Church Father of the fourth century. 
By the application of certain chemical reagents and with the use 
of the ultraviolet-ray lamp, scholars have been able to read 
much of the almost obliterated underwriting, although the task 
of deciphering it is most trying to the eyes. 

In a.d. 692 the council in Trullo (also called the Quinisext 
Council) issued a canon (no. 68) condemning the practice of 
using parchment from manuscripts of the Scriptures for other 
purposes. Despite the canon and the penalty of excommunica- 
tion for one year, the practice must have continued, for of the 
250 uncial manuscripts of the New Testament known today, 
52 are palimpsests. 1 

Instead of writing on the lines, as we do today, ancient scribes 
would ordinarily write the Greek letters pendent from the lines, 
that is, hanging underneath the lines. Usually no spaces were 

1 Besides codex Ephraemi they are the following: P e (= 024), P apr (= 025), 
Q.( = 026), R (= 027), Z (= 035), 3 (= 040), 048, 062, 064, 065, 066, 067, 068, 
072, 078, 079, 086, 088, 093, 094, 096, 097, 098, 0103, 0104, 01 16, 0120, 0130, 
0132, 0133,0134, 0135, 0158, 0159,0161, 0168,0196, oig7, 0208, 02og, 0225, 0229, 
0233, 0240, 0245, 0246, 0247, 0248, 0249, and 0250. See below, pp. 261-2. 

The Making of Ancient Books 13 

left between words or sentences (this kind of writing is called 
scriptio continua), and until about the eighth century punctuation 
was used only sporadically. 1 At times, of course, the meaning of 
a sentence would be ambiguous because the division into words 
was uncertain. In English, for example, godisnowhere will be 
read with totally different meanings by an atheist and by a 
theist ('God is nowhere' and 'God is now here'). It must not be 
thought, however, that such ambiguities occur very often in 
Greek. 2 In that language it is the rule, with very few exceptions, 
that native Greek words can end only in a vowel (or a diph- 
thong) or in one of three consonants, v, p, and s. Furthermore, 
it should not be supposed that scriptio continua presented excep- 
tional difficulties in reading, for apparently it was customary 
in antiquity to read aloud, even when one was alone. 3 Thus, 
despite the absence of spaces between words, by pronouncing to 
oneself what wa& read, syllable by syllable, one soon became 
used to reading scriptio continua.* 

Christian scribes developed a system of contractions for 

1 Word-division, however, is occasionally found in school and liturgical texts, 
and scattered examples of punctuation, by point or spacing or a combination of 
both, are preserved in papyri from the third century B.C. onward. 

2 Examples in the New Testament include the following. In Markx. 40 according 
to most editors Jesus says, '. . . but it is for those for whom it has been prepared' 
(<iAA' 0*5 ijTot/xaarai) . This can also be read dAAotj ij-rot/^aaTat, which means 'it has 
been prepared for others'. In Rom. vii. 14 otSa/xev may be divided ofSa iUv. In 
I Tim. iii. 16 /cat ofioXoyovfievws fitya ccjtiV . . . may be taken as /cat 6fio\oyovfj.ev 
a>s i**ya etjTtV. 

3 Besides scattered evidence from classical antiquity (collected by Josef Balogh, 
'Voces paginarum', Philologus, lxxxii [1927], pp. 84-109, 202-31; also published 
separately), the statement in Acts viii. 30 that Philip 'heard' the Ethiopian treasurer 
reading from Isaiah the prophet implies that he had been reading aloud to himself. 
Compare also the close of 2 Maccabees: 'Here I will end my story. If it is well told 
and to the point, that is what I myself desired ; if it is poorly done and mediocre, 
that was the best I could do. For just as it is harmful to drink wine alone, or, again, 
to drink water alone, while wine mixed with water is sweet and delicious and 
enhances one's enjoyment, so also the style of the story delights the ears of those who 
read the work' (xv. 37-39). See also G. L. Hendrickson, 'Ancient Reading', Classical 
Journal, xxv ( 1 929) , pp. 1 82-96 ; H. J. Chaytor, 'The Medieval Reader and Textual 
Criticism' , Bulletin of the John Rylands Library , xxvi (1941-2), pp. 49-56; and Eugene 
S. McCartney, 'Notes on Reading and Praying Audibly', Classical Philology, lxiii 
( 1 948) , pp. 1 84-7. For a discussion of reading aloud Kara StaoToA^v, see W. G. Ruther- 
ford, A Chapter in the History of Annotation (= Scholia Aristophanica, vol. iii, London, 
J 9°5)> PP- '68 ff. See below, p. 262. 

4 The experience of Hermas, who says he copied a little scroll of heavenly origin 
'letter by letter, for I could not make out the syllables' {Vision, 11. i. 4), suggests that 
the normal method of copying books was by syllables. 

14 The Making of Ancient Books 

certain 'sacred' words. These nomina sacra, as they are called 
today, include such frequently occurring nouns as 9e6s, Kvpios, 
'Irjoovs, Xpioros, and vlos (which were contracted by writing 
only the first and last letters) ; irve€jj,a, Jatn'S, oTavpos, and p-rjr^p 
(contracted by writing only the first two and the last letters) ; 
TTarrjp, 'IoparjX, and oiorrjp (of which the first and the last two 
letters were written) ; and avOpamos, 'IepovaaA-rjp,, and ovpavos (of 
which the first and last syllables were written). In order to draw 
the reader's attention to the presence of a nomen sacrum the scribe 
would place a horizontal line above the contraction. 1 

In the earlier ages of the Church, Biblical manuscripts were 
produced by individual Christians who wished to provide for 
themselves or for local congregations copies of one or more 
books of the New Testament. Because the number of Christians 
increased rapidly during the first centuries, many additional 
copies of the Scriptures were sought by new converts and new 
churches. As a result, speed of production sometimes outran 
accuracy of execution. Furthermore, in preparing translations 
or versions for persons who knew no Greek, it occurred more 
than once (as Augustine complained) that 'anyone who hap- 
pened to gain possession of a Greek manuscript and who 
imagined that he had some facility in both Latin and Greek, 
however slight that might be, dared to make a translation' (De 
doctr. Chr. n. xi. 16). 

When, however, in the fourth century Christianity received 
official sanction from the State, it became more usual for com- 
mercial book manufacturers, or scriptoria, to produce copies of 
the books of the New Testament. Sitting in the workroom of a 
scriptorium, several trained scribes, Christian or non-Christian, 
each equipped with parchment, pens, and ink, would write a 
copy of the book being reproduced as the reader, or lector, 
slowly read aloud the text of the exemplar. 2 In this way as many 

' The standard work of Ludwig Traube, Nomina Sacra: Versuch einer Geschichte der 
christlichen Kiirzung (Munich, 1907), is now supplemented by the additional data 
collected by A. H. R. E. Paap, Nomina Sacra in the Creek Papyri of the First Five 
Centuries A.D.: the Sources and Some Deductions (Leiden, 1959). Traube had fewer than 
forty Greek papyri available ; Paap cites evidence from 42 1 papyri of the first five 
centuries of the Christian era. According to C. H. Roberts, the origin of the dis- 
tinctively Christian custom of contracting the nomina sacra lies in the first century 
(The [London] Times Literary Supplement, 10 Mar. 1961, p. 160). 

2 See especially T. C. Skeat, 'The Use of Dictation in Ancient Book-Production', 
Proceedings of the British Academy, xlii (1956), pp. 179-208. 

The Making of Ancient Books 15 

copies could be produced simultaneously as scribes were working 
in the scriptorium. It is easy to understand how in such a method 
of reproduction errors of transcription would almost inevitably 
occur. Sometimes the scribe would be momentarily inattentive 
or, because of a cough or other noise, would not clearly hear the 
lector. Furthermore, when the lector read aloud a word which 
could be spelled in different ways (as in English, for example, 
the words 'great' and 'grate', or 'there' and 'their'), the scribe 
would have to determine which word belonged in that parti- 
cular context — and sometimes he wrote down the wrong word. 
(For examples of such mistakes, see pp. 190-2 below.) 

In order to ensure greater accuracy, books produced in scrip- 
toria were commonly checked over by a corrector (SiopOcoTrjs) 
specially trained to rectify mistakes in copying. His annotations 
in the manuscript can usually be detected today from differences 
in styles of handwriting or tints of ink. 

Scribes who were hired by a scriptorium to do a certain piece 
of work would be paid in accord with the number of lines which 
they wrote. The standard length of line was originally a line of 
poetry, either a Homeric hexameter or an iambic trimeter. 
When prose works were copied, a line called a stichos, having 
sixteen (or sometimes fifteen) syllables, was frequently used as 
a measure for determining the market price of a manuscript. 
A price-fixing edict issued in a.d. 301 by the Emperor Diocletian 
set the wages of scribes at the rate of 25 denarii for 100 lines 
in writing of the first quality, and of 20 denarii for the same 
number of lines in writing of the second quality (what the differ- 
ence was between the two qualities of writing is not mentioned) . ' 
According to the computation of Rendel Harris, the cost of 
producing one complete Bible, such as codex Sinaiticus, would 
have come to about 30,000 denarii — a sizeable sum notwith- 
standing steadily rising inflation. 2 

The application of stichometric reckoning served also as a 

1 De pretiis return venalium ('Concerning the Prices of Things Sold'), vii. 39 f., 
published in Corpus Inscriptionum Latinarum, iii. 831; see E. R. Graser in Tenney 
Frank's An Economic Survey of Ancient Rome, vol. v (Baltimore, 1940), p. 342. 

1 See Rendel Harris, New Testament Autographs, being a Supplement to the 
American Journal of Philology, no. 12 (Baltimore, 1882), p. 23. It is difficult to estimate 
the exact equivalent in modern currency. Some measure of comparison may be had, 
however, when it is known that in the preceding century under Caracalla (21 1-17) 
a legionary was paid a stipendium of 750 denarii per year in addition to his main- 

1 6 The Making of Ancient Books 

rough and ready check on the general accuracy of a manuscript, 
for obviously a document which was short of the total number 
of stichoi was a defective copy. On the other hand, such cal- 
culations were far from being foolproof safeguards to the purity 
of the text, for only longer interpolations or omissions would be 
likely to be disclosed by counting stichoi. In manuscripts of the 
Gospels which supply stichometric information, the most fre- 
quently appearing statistics are the round numbers 2,600 for 
Matthew, 1,600 for Mark, 2,800 for Luke, and 2,300 for John. 
More precise figures, found in several manuscripts, are 2,560, 
1,616, 2,750, and 2,024 respectively for the four Gospels, which 
imply, for example, the presence of xvi. 9-20 in Mark and the 
absence of vii. 53-viii. n in John. 

Later, during the Byzantine period, copies of books were 
produced by monks. In monasteries there was much less 
pressure than in a commercial scriptorium to produce many 
copies at one time, and so, instead of writing at the dictation of 
a lector, individual monks, often working separately in their 
cells, would prepare copies of the Scriptures or other books either 
for themselves or for some benefactor to the monastery. Such a 
method of multiplying copies was not open to the same kinds 
of errors involved in the dictation method. But another set of 
circumstances operated to make absolute accuracy difficult to 
secure. The act of copying entails four fundamental operations: 
( 1 ) the reading to oneself (in antiquity no doubt reading half- 
aloud) of a line or a clause of the text to be copied, (2) the 
retaining of this material in one's memory, (3) the dictating 
of this material to oneself (either silently or half-aloud) , and 
(4) the movement of the hand in executing the copy. Though 
several of these steps are executed almost simultaneously, there 
was enough opportunity for the mind of a weary or half-awake 
scribe to play tricks that resulted in committing the most atro- 
cious blunders (for examples, see pp. 193-5 below). 

Besides various psychological causes of errors, physiological 
and external causes also conspired to make absolute accuracy 
extremely difficult to attain. One must bear in mind that the act 
of copying was in itself arduous and fatiguing, both because of 
the effort of sustained attention which it demanded as well as 
because of the cramped position of various muscles of the body. 
Though it seems strange to us today, in antiquity it was not 

The Making of Ancient Books 17 

customary to sit at a table or a desk while writing. Both literary 1 
and artistic 2 evidence suggests that until the early Middle Ages 
it was customary for scribes either to stand (while making rela- 
tively brief notes) , or to sit on a stool or bench (or even on the 
ground) , holding their scroll or codex on their knees (see Plates I 
and VII). 3 It goes without saying that such a posture was more 
tiring than sitting at a desk or writing-table — though the latter 
must have been tiring enough to scribes thus occupied six hours 
a day 4 month after month. 

Something of the drudgery of copying manuscripts can be 
learned from the colophons, or notes, which scribes not in- 
frequently placed at the close of their books. A typical colophon 
found in many non-Biblical manuscripts reveals in no uncertain 
terms what every scribe experienced: 'He who does not know 
how to write supposes it to be no labour ; but though only three 
fingers write, the whole body labours.' A traditional formula 
appearing at the close of many manuscripts describes the physio- 
logical effects of prolonged labour at copying : 'Writing bows 

1 An interesting piece of literary evidence is found in a colophon, dating from 
about the third century A.D., attached to a papyrus scroll containing portions of 
the third and fourth books of the Iliad (edited by H.J. M. Milne, Catalogue of the 
Literary Papyri in the British Museum [London, 1927], pp. 21-22). According to the 
reading proposed by Wifstrand, the first two lines of the colophon are as follows : 
'Eyui Kopcovis elfit ypap.p,arojv <fiv\a£' KaXafios p-' eypaifie 8e£ia X^P Ka * L yovv, i.e. be- 
cause the scribe held the sheet of papyrus on his lap while he wrote, it could be said 
that the knee as well as the stylus and the right hand co-operated in producing what 
was written ; cf. Albert Wifstrand, 'Ein metrischer Kolophon in einem Homerus- 
papyrus', Hermes, lxviii (1933), pp. 468-72. 

2 For a variety of artistic evidence bearing on the posture of scribes while writ- 
ing, see the plates in A. M. Friend, Jr., 'The Portraits of the Evangelists in Greek 
and Latin Manuscripts', Art Studies, v (1927), pp. 1 15-47, and vii (1929), pp. 3-29; 
W. H. P. Hatch, Greek and Syrian Miniatures in Jerusalem (Cambridge, Massa- 
chusetts, 1 931); and B. M. Metzger, 'When Did Scribes Begin to Use Writing 
Desks?', in Akten des XL internationalen Byzantinisten-Kongresses, 1958 (Munich, i960), 
pp. 355-62. For a discussion of a miniature in the Rossano Gospels depicting the 
Trial of Jesus (see Plate VII), which suggests that it was customary for stenographers 
in a law court to stand while writing, even though a table was available, see 
William C. Loerke, 'The Miniatures of the Trial in the Rossano Gospels', Art 
Bulletin, xliii (1961), pp. 171-95. See below, p. 262. 

3 The so-called writing-tables found at Qumran, which have been built up by 
the archaeologists to the height of tables today, were originally but 20 inches high, 
too low to serve as writing-desks ; see the present writer's discussion, 'The Furniture 
of the Scriptorium at Qumran', Revue de Qumran, i (1958— g), pp. 509-15. 

4 See the anonymous complaint, dating perhaps from the ninth century, quoted 
by Falconer Madan in his Books in Manuscript (London, 1893), p. 37: 'Ardua 
scriptorum prae cunctis artibus ars est : | Difficilis labor est, durus quoque flectere 
colla, I Et membranas bis ternas sulcare per horas.' 

1 8 The Making of Ancient Books 

one's back, thrusts the ribs into one's stomach, and fosters a 
general debility of the body.' In an Armenian manuscript of the 
Gospels a colophon complains that a heavy snowstorm was 
raging outside and that the scribe's ink froze, his hand became 
numb, and the pen fell from his fingers! It is not surprising that 
a frequently recurring colophon in manuscripts of many kinds 
is the following comparison: 'As travellers rejoice to see their 
home country, so also is the end of a book to those who toil [in 
writing].' Other manuscripts close with an expression of grati- 
tude: 'The end of the book; thanks be to God!' 

Upon more mature reflection, however, many scribes doubt- 
less judged the rewards of copying the Scriptures to outweigh 
the discomforts they experienced during the long hours of writ- 
ing. Thus Cassiodorus, that remarkable rhetorician-philosopher 
and Prime Minister to the Ostrogothic princes of Italy, who later 
became a monk and founded the monastery of Vivarium, noted for 
its school of Latin palaeography, dilates upon the spiritual 
compensations gained by the faithful scribe : 

By reading the divine Scriptures [the scribe] wholesomely instructs 
his own mind, and by copying the precepts of the Lord he spreads 
them far and wide. What happy application, what praiseworthy 
industry, to preach unto men by means of the hand, to untie the 
tongue by means of the fingers, to bring quiet salvation to mortals, 
and to fight the Devil's insidious wiles with pen and ink! For every 
word of the Lord written by the scribe is a wound inflicted on Satan. 
And so, though seated in one spot, the scribe traverses diverse lands 
through the dissemination of what he has written. . . . Man multi- 
plies the heavenly words, and in a certain metaphorical sense, if I 
may dare so to speak, three fingers are made to express the utterances 
of the Holy Trinity. O sight glorious to those who contemplate it 
carefully! The fast-travelling reed-pen writes down the holy words 
and thus avenges the malice of the Wicked One, who caused a reed 
to be used to smite the head of the Lord during his Passion. 1 

In view of the difficulties involved in transcribing ancient 
books, it is the more remarkable how high was the degree of 
achievement of most scribes. The fact is that in most manuscripts 
the size of the letters and the ductus of the script remain sur- 
prisingly uniform throughout even a lengthy document. 

1 Cassiodori Scnatoris Institutiones, edited from the manuscripts by R. A. B. Mynors 
(Oxford, 1937), 1. xxx. I. 

The Making of Amient Books 19 

In order to secure a high degree of efficiency and accuracy, 
certain rules pertaining to the work of scribes were developed 
and enforced in monastic scriptoria. The following are examples 
of such regulations prepared for the renowned monastery of 
the Studium at Constantinople. About a.d. 800 the abbot of 
this monastery, Theodore the Studite, who was himself highly 
skilled in writing an elegant Greek hand, included in his rules 
for the monastery severe punishments for monks who were not 
careful in copying manuscripts. 1 A diet of bread and water was 
the penalty set for the scribe who became so much interested 
in the subject-matter of what he was copying that he neglected 
his task of copying. Monks had to keep their parchment leaves 
neat and clean, on penalty of 1 30 penances. If anyone should take 
without permission another's quaternion (that is, the ruled and 
folded sheets of parchment) , fifty penances were prescribed. If 
anyone should make more glue than he could use at one time, 
and it should harden, he must do fifty penances. If a scribe broke 
his pen in a fit of temper (perhaps after having made some acci- 
dental blunder near the close of an otherwise perfectly copied 
sheet), he had to do thirty penances. 

Added note on colophons 

Besides the several colophons quoted above, which directly 
or indirectly witness to the difficulties involved in transcribing a 
book, there are many other kinds of colophons. Some provide 
the name of the scribe, and occasionally the place and date of 
writing. Obviously such information is of the greatest value to 
the palaeographer in tracing the background and family relation- 
ships of manuscripts. 2 

Some colophons are in the form of a blessing or prayer, or an 
invitation to the reader to offer such a prayer; e.g. 'Whoever 
says, "God bless the soul of the scribe", God will bless his soul.' 
The following prayer is found at the close of a Psalter copied 
in the year 862 : 

1 The text of these rules is in Migne, Patrologia Graeca, xcix. 1 739 f. For a 
modern biography of Theodore, see Alice Gardner's Theodore of Studium, his Life 
and Times (London, 1905). 

1 For a list of the names of Greek scribes, see Marie Vogel and Viktor Gardthausen, 
Die griechischen Schreiber des Mittelalters und der Renaissance (Leipzig, 1909). 

20 The Making of Ancient Books 

eAeos T(S ypaifiavri, Kupte, 
oo<f>la rots avayivwoKovai, 
X<ty>iS rots OLKOVOVai, 
awrrjpia tois kcktt)ixcvois' OLp.tJv. 

(Mercy be to him who wrote, O Lord, wisdom to those who read, 
grace to those who hear, salvation to those who own [this codex]. 

An extended prayer at the end of a Coptic-Arabic manuscript 
of the Gospels contains the following : 

. . . O reader, in spiritual love forgive me, and pardon the daring 
of him who wrote, and turn his errors into some mystic good. . . . 
There is no scribe who will not pass away, but what his hands have 
written will remain for ever. Write nothing with thy hand but that 
which thou wilt be pleased to see at the resurrection. . . . May the 
Lord God Jesus Christ cause this holy copy to avail for the saving 
of the soul of the wretched man who wrote it. 1 

In some manuscripts one finds curse colophons, which were 
intended to serve as the equivalent of modern theft insurance. 
For example, in a twelfth- or thirteenth-century Greek lection- 
ary of the Gospels, now in the library of Princeton Theological 
Seminary (see Fig. 2), there is a colophon stating that the vol- 
ume was donated to the church of St. Saba at Alexandria; the 
colophon continues : 'No one therefore has authority from God 
to take it away under any condition, and whoever transgresses 
this will be under the wrath of the eternal Word of God, whose 
power is great. Gregory, Patriarch by the grace of God, wrote 
this.' 2 

Less formal are the conversational jottings which occasionally 
stand at the close of a manuscript or in the margins of folios 
throughout a document. Though scribes were forbidden to talk 
to one another in the scriptorium, the more irrepressible found 
devious ways to communicate with each other. One such means 
was to jot remarks on the margin of the page being transcribed 
and to show it to one's neighbour. The margins of a ninth-century 
Latin manuscript of Cassiodorus' commentary on the Psalms 

1 [George Horner,] The Coptic Version of the New Testament in the Northern Dialed, 
otherwise called Memphitic and Bohairic, i (Oxford, 1898), pp. cxlvi f. 

2 For other curse colophons, see Lawrence S. Thomson's interesting article, 'A 
Cursory Survey of Maledictions', Bulletin of the Mew Tork Public Library, lvi (1952), 
PP- 55-74- 

The Making of Ancient Books 21 

contain a variety of commonplace remarks written in Irish. 
For example: 'It is cold today.' 'That is natural; it is winter.' 
'The lamp gives a bad light.' 'It is time for us to begin to do 
some work.' 'Well, this vellum is certainly heavy!' 'Well, I call 
this vellum thin!' 'I feel quite dull today; I don't know what's 
wrong with me.' 1 

How did it happen that the head of the scriptorium allowed 
his monks to disfigure a manuscript with such trivialities? One 
may perhaps conjecture that the manuscript was written in a 
continental monastery where the authorities knew no Irish and 
therefore the scribes from Ireland felt they could play pranks 
with impunity. When asked what he had written the scribe 
might point to some pious sentences in Latin in the top margins 
of preceding pages and say, 'Merely the Irish equivalents of 
sentences like these!' 2 

In order to ensure accuracy in transcription, authors would 
sometimes add at the close of their literary works an adjuration 
directed to future copyists. So, for example, Irenaeus attached to 
the close of his treatise On the Ogdoad the following note : 

I adjure you who shall copy out this book, by our Lord Jesus 
Christ and by his glorious advent when he comes to judge the living 
and the dead, that you compare what you transcribe, and correct it 
carefully against this manuscript from which you copy; and also that 
you transcribe this adjuration and insert it in the copy. 3 

IV. 'helps for readers' in new testament 


Many manuscripts of the New Testament are provided with 
a variety of what may be called 'helps for readers', that is, aids to 
assist in the private as well as public reading of the Scriptures. 
This material originated in various places and at various times. 

1 The Irish text, with a German translation, is given by Kuno Meyer in his 
article, 'Neu aufgefundene altirische Glossen', £eitschrift fur celtische Philologie, viii 
(191 2), pp. 173-7. For many other colophons and notes in Greek and Latin manu- 
scripts, see W. Wattenbach, Das Schriftwesen im Mittelalter, 3te Aufl. (Leipzig, 1896), 
pp. 491-534, and Viktor Gardthausen, Griechische Palaeographie, 2te Aufl., ii (Leip- 
zig. '^.PP- 424 ff- 

2 For other instances of conversation written in the margins of manuscripts, see 
W. M. Lindsay, Early Irish Minuscule Script (Oxford, 1910), p. 42. 

3 Apud Eusebius, Hist. Eccl. v. xx. 2. Compare the warning given in Rev. xxii. 18, 
and see W. C. van Unnik, 'De la regie M-qn -npoaBttvai iir/rt a$t\tiv dans l'histoire 
du canon', Vigiliae Christianae, iii (1949), pp. 1-36. 

828158 C 

22 The Making of Ancient Books 

It was handed on from generation to generation, and, as would 
be expected, it grew in volume with the passage of time. The fol- 
lowing is an enumeration of some of these aids found in Greek 
manuscripts. 1 


The oldest system of capitulation which is known to us is that 
preserved in the margins of codex Vaticanus (B) of the fourth 
century. In this manuscript there are 1 70 sections in Matthew, 
62 in Mark, 152 in Luke, and 50 in John. Another system of 
chapter divisions is found in codex Alexandrinus (A) of the 
fifth century as well as in most other Greek manuscripts. Accord- 
ing to this capitulation in Matthew there are 68 Ke<f>d\ata, in 
Mark 48, in Luke 83, and in John 18. It is to be noted that in 
no case does the first chapter stand at the beginning of a book, 
probably because of the custom of scribes to refer to the open- 
ing section of a book as the tt/dooiJluov or preface. Thus Ke<f>. a' of 
Mark begins at Mark i. 23. 

For the Book of Acts several systems of chapter divisions are 
current in the manuscripts. Codex Vaticanus has two sets of 
capitulation, one of thirty-six chapters and the other of sixty- 
nine chapters. According to the opinion of Hatch, the chapter 
numbers of the former system were inserted in the margin of the 
manuscript by a very early hand — perhaps by the StopOajTr/s or 
possibly by the scribe himself — and the other system of chapter 
numbers was added somewhat later by another scribe. 2 In codex 
Sinaiticus of the fourth century the system of sixty-nine chap- 
ters Was added by someone to the first part of Acts (chaps, i— 
xv) , but for some unknown reason it was not continued to the 
end of the book. 

Most other Greek manuscripts of Acts have a system of forty 
Ke<f>dXaia. In some manuscripts the division of Acts into sections 
was carried still further, twenty-four of the forty chapters be- 
ing subdivided into smaller sections (imoSiatpeoeis) . There were 
forty-eight such smaller subdivisions, making a total of eighty- 

1 For a discussion of the accessories in Latin manuscripts of the Bible, see 
[Donatien De Bruyne,] Sommaires, divisions et rubriques de la Bible latine (Namur, : 9 1 4) , 
Prefaces de la Bible latine (Namur, 1920), and Patrick McGurk, Latin Gospel Books 
from A.D. 400 to A.D. 800 (Paris and Brussels, 1961). 

2 W. H. P. Hatch, 'Euthalius', The Twentieth Century Encyclopedia of Religious 
Knowledge, i (Grand Rapids, 1955), p. 400. 

The Making of Ancient Books 23 

eight K€<fxiXaia and imoSicupeoeis. It was inevitable that the 
distinction between the larger and smaller sections would be 
confused, and in some manuscripts they are numbered con- 
secutively throughout the book. 

Both the Pauline and the Catholic Epistles were also divided 
into chapters, and many of these were subdivided into smaller 
sections. 1 Codex Vaticanus has two sets of chapter divisions for 
the Epistles, an earlier and a later system. In the Pauline 
Epistles the earlier division enumerates the chapters consecu- 
tively throughout the whole corpus (for the clue which this 
provides regarding an ancestor of Vaticanus, see p. 48 below). 

The Book of Revelation was supplied with a highly artificial 
system of divisions. At the latter part of the sixth century 
Archbishop Andrew of Caesarea in Cappadocia wrote a com- 
mentary on the book which gives a 'spiritual' exegesis. Instead 
of asking what material was in the book, and into how many 
parts it could most appropriately be divided, he divided the 
book into twenty-four Xoyoi, or discourses, because of the 
twenty-four elders sitting on thrones about the throne of God 
(Rev. iv. 4). He further reflected that the nature of each of the 
twenty-four elders was tripartite (aw/xa, ij/vxrj, and irvev^a), and 
therefore divided each Xoyos into three Ke<f>d\aia, making a total 
of seventy-two chapters for the entire book. 


Each of the Ke<f>dXaia of the system found in codex Alexan- 
drinus and in most other later manuscripts is provided with a 
tiVAo?. This is a summary-heading placed in the margin and 
describing the contents of the chapter (see Plate VIII). These 
titles customarily begin with the word 'about' or 'concerning', 
and are not infrequently written with red ink. Thus Ke<f>. a of 
John, which begins at ii. 1, has the title, 'Concerning the mar- 
riage at Cana' {irepl tov iv Kava ydpav) . 2 All of the titXoi for one 
book are frequently drawn up in a list and placed before that 
book as a summary-outline of what follows. 

1 For information regarding the number of each of these divisions in each of the 

Epistles, see W. H. P. Hatch, Facsimiles and Descriptions of Minuscule Manuscripts of 
the New Testament (Cambridge, Massachusetts, 1951), p. 25. See below, p. 262. 

2 For a complete list of titAoi, see H. von Soden, Die Schriften des Neuen Testa- 
ments in inter altesten erreichbaren Textgestalt, 1. i (Berlin, 1902), pp. 405 ff. 

24 The Making of Ancient Books 


An ingenious system was devised by Eusebius of Caesarea to 
aid one in locating parallel passages in the Gospels. Apparently 
his system was found to be highly useful, for it appears in a 
great number of manuscripts of the Gospels in Greek, as well 
as in Latin, Syriac, Coptic, Gothic, Armenian, and other 

The synopsis or harmony was prepared as follows. Each 
Gospel was divided into longer or shorter sections, depending 
upon the relation of each section to one or more parallels in the 
other Gospels. These sections were numbered consecutively 
throughout each Gospel (there are 355 in Matthew, 233 in 
Mark, 342 in Luke, and 232 in John). Then Eusebius prepared 
ten tables or canons (tcavoves) , the first containing references by 
numerals to parallel passages found in all four Gospels; the 
second to passages common to Matthew, Mark, and Luke ; the 
third to passages common to Matthew, Luke, and John ; and 
so on, until almost all possible combinations of Gospels were 
exhausted. 1 The final table gives references to matter pecu- 
liar to each Gospel alone. These tables of numerals, written 
out in columns, customarily occupy the opening pages of Gos- 
pel manuscripts. 2 Then, in the margin of the Gospel text of 
the manuscript, alongside or beneath the numeral of the con- 
secutive sections, was written the numeral of the canon table 
in which that section could be found. For example, if while 
reading the Gospel according to John one came upon the state- 
ment, 'For Jesus himself testified that a prophet has no honour 
in his own country' (iv. 44) , and wished to consult the parallels, 
he would see in the margin the numerals £ (=^£). By turning to 
the first canon table and running his eye down the column of 
numerals referring to sections in John, he would find 35. In the 
horizontal line opposite this numeral he would then find the 
numeral 142 standing in the column of sections of Matthew, 51 
in Mark, and 2 1 in Luke. Since, as was said above, the sections 

1 There is no table of references to sections in Mark, Luke, and John ; nor of 
references to sections in Mark and John. 

2 The bare lists of numerals soon attracted the attention of artists, who orna- 
mented the spaces between with decorative columns, arches, architraves, birds, 
flowers, &c. ; see Plate Xlli, and Carl Nordenfalk, Die spatantiken Kanontafeln : 
Kunstgeschichtliche Studien iiber die Eusebianische Evangelien-Konkordanz in den vier ersten 
Jahrhunderten ihrer Geschichte, 2 vols. (Goteborg, 1938). See below, p. 262. 

The Making of Ancient Books 25 

in each Gospel are numbered consecutively, it is an easy matter 
to find in the other three Gospels each of the sections that con- 
tain the parallels to the statement in John. 

For the added convenience of the user, in some manuscripts 
the numerals referring to the sections which are parallel to the 
passage on any given page are provided in the lower margin of 
the page (see Plates VIII and Xllb), so that one can turn at 
once to the appropriate sections in the other Gospels. 

Eusebius explained his ingenious system in a letter to a fellow 
Christian named Carpianus, and a copy of this letter appears 
with the canon tables on the opening folios of many Gospel 
manuscripts. 1 Several twentieth-century editions of the Greek 
Testament are provided with the Eusebian canon tables and 
numbers, which thus continue to be of service to present-day 
readers of the Gospels. 


The hypothesis (vnodeois, Latin argumentum) is a prologue or 
brief introduction to a book, supplying the reader with a cer- 
tain amount of information concerning the author, content, 
and circumstances of composition of the particular book. The 
form and content of such hypotheses are often conventional 
and stereotyped. In some manuscripts the hypotheses for the 
Gospels are ascribed to Eusebius, but more often they are 
anonymous. Anti-Marcionite Gospel prologues are extant in 
Latin manuscripts from the fifth to the tenth century. The 
Marcionite prologues to the (ten) Pauline Epistles were taken 
over practically unaltered by the Roman Catholic Church for 
the Latin Vulgate. 2 

A longer statement of traditional information concerning the 
life of each evangelist (called his filos) sometimes appears with 
the hypothesis. The lives are attributed to an otherwise un- 
known Dorotheus of Tyre or to Sophronius, the Patriarch of 
Jerusalem in the first half of the seventh century. 

Several different prologues which define the word 'gos- 
pel' and provide general information about all four Gospels 

1 An English translation of Eusebius' letter was published by Harold H. Oliver 
in Novum Testamentum, iii (1959), pp. 138-45. See below, p. 262. 

2 For a survey of recent investigations concerning the Marcionite prologues, see 
B. M. Metzger in The Text, Canon, and Principal Versions of the Bible, ed. by E. E. 
Flack and B. M. Metzger (Grand Rapids, 1956), pp. 24 ff. 

26 The Making of Ancient Books 

collectively occur in various manuscripts. Besides lists of the 
twelve apostles, the traditional names of the seventy[-two] dis- 
ciples in Luke x. i ff. are given on the authority of Dorotheus 
and Epiphanius. 

For the Book of Acts several different hypotheses are current ; 
some are anonymous and others are taken from Chrysostom's 
commentary and homilies on Acts. For Acts and the Epistles a 
considerable apparatus of auxiliary materials circulated under 
the name of Euthalius or Evagrius. 1 Besides chapter divisions 
and hypotheses, these included a lengthy sketch of the life, 
writings, and chronology of the apostle Paul ; a brief statement 
of the martyrdom of Paul ; a table of Old Testament quotations 
in the Epistles; a list of places at which the Epistles were thought 
to be written ; and a list of the names associated with Paul's in 
the headings to the Epistles. How much of this supplementary 
material was drawn up by Euthalius and how much was added 
later is not known. 


In the oldest manuscripts of the New Testament the titles of 
the several books are short and simple, e.g. KATA MA00AION 
or TIPOC PQMAIOYC. But in later centuries these titles became 
longer and more complex (see p. 205 below). 

The subscriptions appended to the end of the books were 
originally (like the titles) brief and simple, merely indicating the 
close of the book. As time passed these became more elaborate, 
and often included traditional information regarding the place 
at which the book was thought to be written and sometimes the 
name of the amanuensis. The King James version includes the 
subscriptions to the Pauline Epistles. 


As was mentioned above, the earliest manuscripts have very 
little punctuation. The Bodmer papyri and the Chester Beatty 

1 Scholarly opinion as to who this Euthalius or Evagrius was and when he lived 
has vacillated widely; see, for example, J. A. Robinson, Euthaliana (Cambridge, 
1895); E. von Dobschiitz, 'Euthaliusstudien', Z e ' tsc hrift fur Kirchengeschichte, xxix 
(1899), pp. 107-54; H. von Soden, op. cit., 1. i (1902), pp. 637-82; C. H. Turner 
in Hastings's Dictionary of the Bible, extra vol. (1904), pp. 524-9; G. Bardy in Sup- 
plement au Dktionnaire de la Bible, ii (1934), cols. 1215-18; G. Zuntz, 'Euthalius 
= Euzoius?', Vigiliae Christianae, vii (1953), pp- 16-22; and W. H. P. Hatch (in 
footnote 2 on p. 22 above). 

The Making of Ancient Books 27 

papyri (see pp. 36-42 below) have only an occasional mark of 
punctuation, 1 as do also the early uncial manuscripts. A diae- 
resis is sometimes placed over an initial iota or upsilon. During 
the sixth and seventh centuries scribes began to use punctua- 
tion marks more liberally, though the sign of interrogation is 
rarely found before the ninth century. Gradually the earlier 
sporadic and somewhat haphazard usage gave way to a fuller 
and more or less consistent style. In addition to the usual marks 
of punctuation a syllable divider was often inserted after non- 
Greek proper names, as in such sequences as that in Matt. i. 2, 
ABPAAM'ErENNHCEN, where it stands to warn the reader 
against dividing it a-braa-me-gen-ne-sen. 


Glosses are brief explanations of difficult words or phrases. 
They were usually written in the margin of manuscripts, though 
occasionally they were placed between the lines. In the latter 
case a Greek manuscript might be 'glossed' throughout with a 
Latin interlinear, and a Latin manuscript with an Anglo-Saxon 

Scholia are interpretative remarks of a teacher placed beside 
the text in order to instruct the reader. When scholia are syste- 
matically developed in order to elucidate continuously the entire 
text, rather than being merely random notes on certain passages, 
the work is called a commentary. Scholia and commentaries are 
sometimes placed in the margins around the Scripture text 
and sometimes interspersed between sections of the Scripture 
text (see Plate IX). In uncial manuscripts the explanatory 
materials are usually in minuscule script, though in codex 
Zacynthius (S 1 ) of the seventh or eighth century the commentary 
is written in uncials. When the text is in minuscules, occasionally 
the scholia are in small uncials. 

Catenae are literally 'chains' of comments extracted from 
older ecclesiastical writers. The identity of the original commen- 
tator is indicated by prefixing the abbreviation of his name, 
though through carelessness this mark of identification is some- 
times missing. 

1 For a list of papyri which have marks of punctuation, see Guilelmus Flock, De 
graecorum interpunctionibus (Diss. Bonn, 1908), pp. 14 ff. 

28 The Making of Ancient Books 

Onomastica are philological aids which purport to give the 
meaning and etymology of proper names. Like so much of 
ancient etymological lore — preserved, for example, throughout 
the works of Philo and in Plato's Cratylus — these explanations 
almost always are arbitrary and fanciful. 


In addition to an ornamental headpiece at the beginning of 
a book and illuminated initials, the Greeks of the Byzantine 
period sought to assist the reader in understanding the signifi- 
cance of the Scriptures by providing pictures of various kinds. 1 
Some are portraits of Christ and of his apostles, and others 
depict scenes or events that are narrated in the New Testament 
(see Plate VII). The portraits of the Evangelists fall into two 
main classes, those in which the figures are standing and those 
in which they are sitting. From a comparison with Hellenistic 
representations of ancient Greek poets and philosophers it ap- 
pears that Christian artists, who had no knowledge of the like- 
nesses of the Evangelists, adopted and adapted familiar portraits 
of pagan authors in contemporary art. According to the 
investigations of A. M. Friend, Jr., 2 all the early Christian 
portraits of the Evangelists go back to two main sets of four 
portraits each: one set was of four philosophers, Plato, Aristotle, 
Zeno, and Epicurus, and the other set was of four playwrights, 
Euripides, Sophocles, Aristophanes, and Menander. 

The earliest New Testament manuscripts which contain 
miniatures are two de luxe copies on purple vellum of the sixth 
century, namely codex Sinopensis (O) and codex Rossanensis 
(Z; see pp. 55 and 59 below). In the course of time custom 
and tradition came to dictate the proper form and colours 
which artists should use in painting the several cycles of Biblical 
scenes and characters. These traditional directives are given in 
the Byzantine Painters' Manual compiled by Dionysius, a monk 

1 For reproductions of representative miniatures, see W. H. P. Hatch, Greek and 
Syrian Miniatures in Jerusalem (Cambridge, Massachusetts, 1931); E.J. Goodspeed, 
D. W. Riddle, and H. R. Willoughby, The Rockefeller McCormick New Testament 
(Chicago, 1932); E. C. Colwell and H. R. Willoughby, The Four Gospels of Kara- 
hissar (Chicago, 1936) ; and H. R. Willoughby and E. C. Colwell, The Elizabeth Day 
McCormick Apocalypse (Chicago, 1940). For a general introduction, see David 
Diringer, The Illuminated Book, its History and Production (New York, 1958). 

2 'The Portraits of the Evangelists in Greek and Latin Manuscripts', Art Studies, 
v ('927). PP- "5-46, and vii (1929), pp. 3-29. 

The Making of Ancient Books 29 

of Fourna d'Agrapha. 1 Unfortunately, New Testament scholars 
have not yet availed themselves of the evidence supplied by 
artistic adornment in the investigation of family relationships of 
Byzantine manuscripts. 2 


The helps mentioned above were designed chiefly to assist 
private readers of Scripture ; the following helps were intended 
to provide aid for the public reading of Scripture lessons in 
services of worship. 

The practice of writing treatises in short lines according to 
the sense antedated its application to Christian writings. Some 
of the orations of Demosthenes and Cicero were transcribed in 
this manner to assist the reader in making the correct inflection 
and the proper pauses. It was also applied to the Septuagint 
Greek text of the poetical books of the Old Testament, 3 and 
when Jerome translated the Prophets into Latin he arranged 
the text colometrically. 4 Each sense-line consisted of a single 
clause (kojAov) or a single phrase (kc^io.). 5 

Several bilingual Greek and Latin manuscripts of the Gospels, 
Acts, and Epistles have the text colometrically arranged; these 

1 Edited by Athanasios Papadopoulos-Kerameus, Manuel d'iconographie chri- 
tienne .. . (St.-Petersbourg, 1909). 

2 For suggestions regarding the relationship between picture criticism and 
textual criticism, see Kurt Weitzmann, Illustrations in Roll and Codex, a Study of the 
Origin and Method of Text Illustration (Princeton, 1947), pp. 182-92. 

3 One of the earliest examples of a portion of the Septuagint arranged in cola is 
the second- (or third-)century Bodleian fragment of the Psalms edited by J. W. B. 
Barns and G. D. Kilpatrick (Proceedings of the British Academy, xliii [1957], 
pp. 229-32). 

4 Jerome, Preface to Isaiah (Migne, Patrologia Latina, xxviii. 825). 

5 In antiquity rhetoricians defined a comma as any combination of words that 
makes a total of not more than eight syllables, while they required from a colon a 
combination of at least nine, though not exceeding sixteen; see James A. Kleist, 
S.J., 'Colometry and the New Testament', Classical Bulletin, iv (1928), pp. 26 f. 
Kleist remarks that 'these figures will not seem arbitrary if we bear in mind that the 
average length of a hexameter is sixteen syllables, and that the average speaker does 
not easily go beyond sixteen syllables without renewing his breath. ... In writing 
colometrically, the one great purpose agreed upon by all ancient rhetoricians is to 
enable the reader to read intelligently, and properly to husband his breath in 
speaking. The all-essential thing is that both colon and comma, when taken by 
themselves, make sense and admit of proper breathing. But essential as this require- 
ment is, it is also a source of arbitrariness of interpretation. How much, after all, or 
how little is required to make sense? How much can be uttered in one breath? 
Here, as elsewhere, quot capita, tot sententiae, and unusquisque in suo sensu abundat.' 
See Plate Via for an example of unusually short lines. 

30 The Making of Ancient Books 

include codex Bezae (D; see Plate V), codex Claromontanus 
(Dp), and codex Coislinianus or Euthalianus (Hp), all dating 
from about the sixth century. According to the recent investiga- 
tions of N. A. Dahl, the original edition from which these copies 
were derived 'must have been a product of ancient book- 
publishing of high standards, in some way connected with the 
traditions of the library at Caesarea'. 1 

10. .NEUMES 

Neumes are Byzantine musical notes which assisted the lector 
in chanting or cantillating the Scripture lesson. They appear 
first in codices of the seventh or eighth century, but whether 
they were contemporary with the text of the manuscript or were 
added later it is difficult to determine. Their form is that of 
hooks, dots, and oblique strokes (see Plate XI), and they are 
usually written with red (or green) ink above the words to be 
sung. 2 


Following the custom of the synagogue, according to which 
portions of the Law and the Prophets were read at divine ser- 
vice each Sabbath day, the Christian Church adopted the 
practice of reading passages from the New Testament books at 
services of worship. A regular system of lessons from the Gospels 
and Epistles was developed, and the custom arose of arranging 
these according to a fixed order of Sundays and other holy days 
of the Christian year. In order to assist the lector in finding the 
beginning {apxrj) and the end (reXos) of the lesson, several of the 
later uncial manuscripts were provided with the abbreviations 
apx and reX, inserted either in the margin or between the lines of 
the text (see Plate VIII). Lection notes, indicating that a given 
passage is to be read on a certain day, were sometimes written 

1 Quoted from a summary of a paper entitled 'Bilingual Editions of the Pauline 
Letters', read by Nils A. Dahl at the annual meeting of the Society of Biblical 
Literature and Exegesis held at St. Louis in 1961. 

2 See especially H. J. W. Tillyard, Handbook of the Middle Byzantine Musical 
Notation {Monumenta Musicae Byzantinae, Subsidia, 1. i; Copenhagen, 1935) ; Carsten 
Hoeg, La Notation ekphonitique (Monumenta Musicae Byzantinae, Subsidia, 1. ii; 
Copenhagen, 1935; and E. G. Wellesz, A History of Byzantine Music and Hymno- 
graphy, 2nd ed. (Oxford, ig6i). See below, p. 262. 

The Making of Ancient Books 31 

in the margin with red ink. A list of all these passages is occasion- 
ally given at the close (or at the beginning) of a codex. 

For the added convenience of the reader, lectionary manu- 
scripts were prepared which present in proper sequence (begin- 
ning with Easter) the text of the several passages of Scripture 
appointed to be read on Sundays, Saturdays, and, in some 
cases, on week-days throughout the year. Such lectionaries are 
called synaxaria (see Plate X) . Another service book is the meno- 
logion, which supplies Scripture lessons for feast days, saints' 
days, and the like, starting with the first of September, the be- 
ginning of the civil year. It is noteworthy that substantially 
the same choice of Scripture passages in lectionary manuscripts 
dating from the seventh or eighth century is still followed by the 
Greek Orthodox Church today. 

Scholars have only recently begun to appreciate the impor- 
tance of lectionary manuscripts in tracing the history of the text 
of the New Testament during the Byzantine period. 1 Inasmuch 
as the form of the citation of the Scriptures in official liturgical 
books always tends to be conservative and almost archaic, 2 
lectionaries are valuable in preserving a type of text that is 
frequently much older than the actual age of the manuscript 
might lead one to suspect. 


It is customary to classify the Greek manuscripts of the 
New Testament into several categories, partly according to the 
material from which they are made, partly according to their 
script, and partly according to the use for which they were 

1 For an introduction to the study of New Testament Greek lectionaries, see 
Ernest C. Colwell and Donald W. Riddle, Prolegomena to the Study of the Lectionary 
Text of the Gospels (Chicago, 1933). See below, p. 262. 

2 Thus, for example, in the Anglican Book of Common Prayer the Psalter and the 
Canticles from Luke are still in the translation of the Great Bible of 1539, despite 
repeated proposals to make them conform to the King James version of 1 6 1 1 . 

3 Prof. Kurt Aland of Munster, who currently assigns official numbers to newly 
discovered Greek manuscripts of the New Testament, graciously supplied the 
present writer (in a letter dated 11 July 1962) with information regarding the 
total number of New Testament papyri, uncials, minuscules, and lectionaries. 
For further information reference may be made to Aland's recent volume, Kurz- 
gefafite Liste der griechischen Handschriften des Neuen Testaments: 1. Gesamtubersicht 
(Berlin, 1963). 

32 The Making of Ancient Books 

At first editors of the New Testament used cumbersome titles 
to designate the Greek manuscripts. These titles were usually 
derived from the names of the owners or the libraries possessing 
the manuscripts. Since no one system was agreed upon by all 
editors, and since manuscripts would change owners and loca- 
tions, it was exceedingly confusing to compare the evidence in 
one critical apparatus with that in another. The first step in the 
direction of standardization of nomenclature was taken by a 
Swiss scholar, Johann Jakob Wettstein, who, in his handsome 
two-volume edition of the Greek Testament published at 
Amsterdam in 1 751-2, employed capital letters to designate 
uncial manuscripts and Arabic numerals to designate minuscule 
manuscripts. The system now in general use was elaborated at 
the end of the nineteenth century by Caspar Rene Gregory, a 
native Philadelphian who, after receiving his theological train- 
ing at Princeton, went to Germany where he became Professor 
of New Testament at the University of Leipzig in 1889. Building 
upon Wettstein's system, Gregory devised several other cate- 
gories of the materials. Thus the manuscripts made of papyrus 
are listed separately from those made of parchment. Each of 
them is commonly referred to by the Gothic or Old English 
letter p followed by a small superior numeral. At the time of this 
writing (1967), a total of eighty-one Greek papyri of the New 
Testament have been assigned official numbers by Gregory and 
his successors. 

Following Wettstein's system, the uncial manuscripts which 
have been known for the longest time are commonly designated 
in a critical apparatus by capital letters of the Latin and Greek 
alphabets, and by one Hebrew letter (X, aleph). Since, however, 
the number of known uncial manuscripts came to exceed the 
number of letters in the Latin, Greek, and Hebrew alphabets 
combined, Gregory assigned to each uncial manuscript an 
Arabic numeral preceded by a zero. Thus far 266 uncial manu- 
scripts have been catalogued. The minuscule manuscripts, 
which, as was mentioned above, are referred to by Arabic 
numbers, now total 2,754. (See below, p. 262.) 

A subsidiary class of Greek manuscripts, involving both uncial 
and minuscule manuscripts (though the latter by far pre- 
dominate in number), is devoted to lectionaries. As was men- 
tioned earlier, these are church reading books containing the 

The Making of Ancient Books 33 

text of selections of the Scriptures appointed to be read on the 
several days of the ecclesiastical and the civil year, compris- 
ing, respectively, the synaxarion and the menologion. Although 
2,135 lectionaries of the Greek New Testament have been cata- 
logued, only comparatively few have been critically studied. 1 In 
the Gregory system of designating manuscripts, lectionaries are 
indicated by the letter 7' followed by an Arabic numeral. Thus 
7' alone designates a Gospel lectionary ; 7 a ' designates a lection- 
ary of the Acts and the Epistles; and 7 +a ' designates a lectionary 
containing lessons from Gospels, Acts, and Epistles. (The Greek 
lectionary contains no lessons from the Book of Revelation.) 

Two other forms of New Testament witnesses may be de- 
scribed with only a few words. Short portions of six New Testa- 
ment books have been preserved on ostraca, or broken pieces of 
pottery used by the poorest people as writing material. Twenty- 
five of these have been catalogued, and are sometimes referred 
to by the Gothic or Old English letter <B followed by a superior 
numeral. Finally, a curious but unimportant source of our know- 
ledge of the Greek text of the New Testament consists of a num- 
ber of talismans, or good-luck charms. These amulets range in 
date from the fourth to the twelfth or thirteenth century and 
are made of vellum, papyrus, potsherd, and wood. The super- 
stitious use of talismans, so prevalent in the ancient world, 
was scarcely less popular among the faithful than among the 
pagans — if we may judge from repeated remonstrances against 
them issued by ecclesiastical authorities. 2 Four of those cata- 
logued contain the Lord' s Prayer and five others include scattered 
verses from other parts of the Old and New Testaments. They 
are sometimes referred to by the letter % followed by a superior 

In evaluating the significance of these statistics of the amount 

1 Scholars at the University of Chicago have been sponsoring the study of this 
long- neglected source of information regarding the text of the New Testament ; see 
the series of monographs published under the general title, 'Studies in the Lection- 
ary Text of the Greek New Testament', begun by E. C. Colwell and D. W. Riddle, 
and now edited by A. Wikgren. See below, pp. 262-3. 

2 Thus, in addition to remonstrances by Eusebius and Augustine, the Synod of 
Laodicea issued a separate canon proscribing the manufacture and use of amulets : 
'. . . and those who wear such we command to be cast out of the Church'. For 
these and other references, see the annotations on a papyrus fever amulet edited by 
the present writer in Papyri in the Princeton University Collections, iii (Princeton, 1 942) , 
pp. 78 f. See below, p. 263. 

34 The Making of Ancient Books 

of Greek evidence for the text of the New Testament, one 
should consider, by way of contrast, the number of manuscripts 
which preserve the text of the ancient classics. Homer's Iliad, 
for example, the 'bible' of the ancient Greeks, is preserved in 
457 P a PY r i> 2 uncial manuscripts, and 188 minuscule manu- 
scripts. 1 Among the tragedians the witnesses to Euripides are 
the most abundant ; his extant works are preserved in 54 papyri 
and 276 parchment manuscripts, almost all of the latter dating 
from, the Byzantine period. 2 

The works of several ancient authors are preserved to us 
by the thinnest possible thread of transmission. For example, 
the compendious history of Rome by Velleius Paterculus sur- 
vived to modern times in only one incomplete manuscript, 
from which the editio princeps was made — and this lone manu- 
script was lost in the seventeenth century after being copied 
by Beatus Rhenanus at Amerbach. Even the Annals of the 
famous historian Tacitus is extant, so far as the first six books 
are concerned, in but a single manuscript, dating from the 
ninth century. In 1870 the only known manuscript of the Epistle 
to Diognetus, an early Christian composition which editors usually 
include in the corpus of Apostolic Fathers, perished in a fire at 
the municipal library in Strasbourg. In contrast with these figures, 
the textual critic of the New Testament is embarrassed by the 
wealth of his material. 3 Furthermore, the work of many an 
ancient author has been preserved only in manuscripts which date 

1 For a list of papyri of the Iliad, see H.J. Mette, 'Neue Homer- Papyri', Revue de 
philologie, xxix (1955), pp. 193-9, 202_ 4, an d Lustrum, i (1956), p. 9, Anm. 1 ; for 
a list of uncial and minuscule manuscripts, see T. W. Allen, Homeri Mas, vol. i 
(Oxford, 1931). See below, p. 263. 

2 For a list of papyri of Euripides, see Roger A. Pack, The Greek and Latin Literary 
Texts from Greco-Roman Egypt (Ann Arbor, 1952), pp. 23 ff. ; for a list of the Byzan- 
tine manuscripts, see Alexander Turyn, The Byzantine Manuscript Tradition of the 
Tragedies of Euripides (= Illinois Studies in Language and Literature, vol. xlv; Urbana, 


3 Lest, however, the wrong impression be conveyed from the statistics given 
above regarding the total number of Greek manuscripts of the New Testament, 
it should be pointed out that most of the papyri are relatively fragmentary and 
that only about fifty manuscripts (of which codex Sinaiticus is the only uncial manu- 
script) contain the entire New Testament. The great majority of the other manu- 
scripts contain the four Gospels. The Book of Revelation is the least well-attested 
part of the New Testament, being preserved in about 300 Greek manuscripts. Of 
this number only ten are uncial manuscripts (namely K, A, C, P, 046, 051, 052, 
0163, 0169, and 0207), and three of these ten comprise only a single leaf each (the 
last three mentioned). 

The Making of Ancient Books 35 

from the Middle Ages (sometimes the late Middle Ages), far 
removed from the time at which he lived and wrote. On the 
contrary, the time between the composition of the books of the 
New Testament and the earliest extant copies is relatively brief. 
Instead of the lapse of a millennium or more, as is the case of not 
a few classical authors, several papyrus manuscripts of portions 
of the New Testament are extant which were copied within a 
century or so after the composition of the original documents. 


Important Witnesses to the 
Text of the New Testament 

Three classes of witnesses are available for ascertaining 
the text of the New Testament ; they are the Greek manu- 
scripts, the ancient translations into other languages, and the 
quotations from the New Testament made by early ecclesiastical 
writers. Something must be said now about certain individual 
witnesses in each of these three classes of evidence. 


Of the approximately 5,000 Greek manuscripts which contain 
all or part of the New Testament, the following are among the 
most important. They are listed here under the usual categories 
of (1) papyri, (2) uncials, and (3) minuscules; within each of 
these groups the sequence is that of the Gregory system of 
numeration. In the descriptions of these manuscripts reference 
is frequently made to various types of text, such as the Alexan- 
drian, the Western, the Caesarean, and the Koine or Byzantine 
forms of text; for information concerning the significance of 
such terminology, see pp. 212-18 below. The name of the edi- 
tor or collator is given for those manuscripts which have been 
published individually; it is to be understood that a more or 
less full conspectus of readings of the other manuscripts men- 
tioned here may be found in the standard apparatus critici. 


Two of the most important collections of papyrus manuscripts 
of the New Testament were acquired in 1 930-1 by Mr. (now 
Sir) Chester Beatty of London and by M. Martin Bodmer of 
Geneva in about 1955-6. The former collection is now in the 

1 For a list of all known Greek papyri of the New Testament, see the Appendix, 
pp. 247-55 below. 

Important Witnesses to the Text 37 

Beatty Museum in a suburb of Dublin and has been edited, with 
introductions and discussions, by Sir Frederic G. Kenyon. 1 

p 45 . The first of the Chester Beatty Biblical papyri, to which 
the siglum p 45 has been assigned, comprises portions of thirty 
leaves of a papyrus book formed by a succession of quires of 
only two leaves. Originally the codex consisted of about 220 
leaves, each measuring about 10 by 8 inches, and contained all 
four Gospels and the Acts. Today Matthew and John are the 
least well preserved, each being represented by only two frag- 
mentary leaves. Six leaves of Mark, seven of Luke, and thirteen 
of Acts remain of these books. Several small fragments of the 
codex, originally comprising a leaf from Matthew, have turned 
up in a collection of papyri at Vienna. 2 

The manuscript is dated by the editor in the first half of the 
third century. The type of New Testament text which it pre- 
serves in Mark is nearer to the Caesarean family than to either 
the Alexandrian or the Western text-types. In the other Gospels 
(where the Caesarean text has not yet been fully established) 
it is also intermediate between Alexandrian and Western. In 
Acts it is decidedly nearer the Alexandrian and has none of the 
major variants characteristic of the Western text of this book, 
though it has some of the minor ones. 

p 46 . The second Chester Beatty Biblical papyrus, designated 
p 46 , comprises eighty-six leaves (all slightly mutilated) of a single- 
quire 3 papyrus codex, measuring originally about 11 by 6^ inches, 
which contained on 104 leaves ten Epistles of Paul in the follow- 
ing order : Romans, Hebrews, 1 and 2 Corinthians, Ephesians, 
Galatians, Philippians, Colossians, 1 and 2 Thessalonians. Slightly 
earlier than p 4 * it dates from about the year 200. Today por- 
tions of Romans and 1 Thessalonians, and 2 Thessalonians 
in its entirety, are lacking. The Pastoral Epistles were prob- 
ably never included in the codex, for there does not appear to 

1 The Chester Beatty Biblical Papyri, Descriptions and Texts . . . (London, 1933-7). 

2 Edited by Hans Gerstinger, 'Ein Fragment des Chester Beatty-Evangelien- 
kodex in der Papyrussammlung der Nationalbibliothek in Wien', Aegyptus, xiii 

('933). PP- 67-72- 

3 Three and possibly four of the Beatty Biblical papyri are single-quire manu- 
scripts. For a discussion of this form of codex, see Campbell Bonner's introduction 
to his Papyrus Codex of the Shepherd of Hernias (Ann Arbor, 1934), pp. 7-12. More 
recently several other examples of single-quire codices have come to light, includ- 
ing the Michigan Gospel of John in Fayyumic Coptic and twelve of the thirteen 
Gnostic manuscripts from Nag Hammadi. 

826158 r, 

38 Important Witnesses to the 

be room for them on the leaves missing at the end. (Since it 
is a single-quire codex, the number of leaves lacking at both 
ends can be computed more or less accurately.) Thirty of the 
eighty-six surviving leaves are in the library at the University 
of Michigan. 1 

It will be observed that, in addition to the reversal of the 
present order of Galatians and Ephesians, the anonymous 
Epistle to the Hebrews is included among the Pauline Epistles, 
which are arranged in a general order of their decreasing 
lengths, p 46 is noteworthy, likewise, in that the doxology to 
Romans, which in many of the earlier manuscripts stands at 
the end of chapter xiv, is here placed at the end of chapter xv 
(see Plate II). 2 In-general the papyrus is closer to the Alexan- 
drian than to the Western type of text. 

p 47 . The third Chester Beatty Biblical papyrus of the New 
Testament, designated p 47 , comprises ten slightly mutilated 
leaves of a codex, measuring about g-| by 5! inches, of the Book 
of Revelation. Of the original codex, estimated to have been 
thirty-two leaves in length, only the middle portion remains, 
containing the text of ix. 10-xvii. 2. It dates from the middle or 
latter part of the third century. In general the text of p 47 agrees 
more often with that of codex Sinaiticus than with any other, 
though it often shows a remarkable independence. 

p 52 . Measuring only i\ by 3^ inches and containing but a 
few verses from the Fourth Gospel (xviii. 31-33, 37-38), this 
papyrus fragment is the oldest copy of any portion of the New 
Testament known to be in existence today. Although it had 
been acquired in Egypt by Bernard P. Grenfell as long ago as 
1920, it remained unnoticed among hundreds of similar shreds 
of papyri until 1934. In that year C. H. Roberts, Fellow of 
St. John's College, Oxford, while sorting over the unpublished 
papyri belonging to the John Rylands Library at Manches- 
ter, recognized that this scrap preserves several sentences from 
John's Gospel. Without waiting to edit the fragment along with 

1 Edited .by Henry A. Sanders, A Third-Century Papyrus Codex of the Epistles oj 
Paul (Ann Arbor, 1935). See below, p. 263. 

2 This papyrus, however, contrary to common opinion, is not alone in plac- 
ing Hebrews immediately following Romans; in six minuscule manuscripts and 
in a Syrian canon composed about a.d. 400 Hebrews occupies this position; see 
W. H. P. Hatch, 'The Position of Hebrews in the Canon of the New Testament', 
Harvard Theological Review, xxix (1936), pp. 133-51- 

Text of the New Testament 39 

others of a miscellaneous nature, he immediately published a 
booklet setting forth a description of the fragment, its text, and 
a discussion of its significance. 1 

On the basis of the style of the script, Roberts dated the 
fragment in the first half of the second century. Though not all 
scholars are convinced that it can be dated within so narrow a 
range, such eminent palaeographers as Sir Frederic G. Kenyon, 
W. Schubart, Sir Harold I. Bell, Adolf Deissmann, Ulrich 
Wilcken, and W. H. P. Hatch have expressed themselves as 
being in agreement with Roberts's judgement. 2 

Although the extent of the verses preserved is so slight, in one 
respect this tiny scrap of papyrus possesses quite as much evi- 
dential value as would the complete codex. Just as Robinson 
Crusoe, seeing but a single footprint in the sand, concluded that 
another human being, with two feet, was present on the island 
with him, so p 52 proves the existence and use of the Fourth 
Gospel during the first half of the second century in a pro- 
vincial town along the Nile, far removed from its traditional 
place of composition (Ephesus in Asia Minor). Had this little 
fragment been known during the middle of the past century, that 
school of New Testament criticism which was inspired by the 
brilliant Tubingen professor, Ferdinand Christian Baur, could 
not have argued that the Fourth Gospel was not composed 
until about the year 1 60. 

p 66 . The most important discoveries of New Testament manu- 
scripts since the purchase of the Chester Beatty papyri are the 
acquisitions made by the Genevan bibliophile and humanist, 
M. Martin Bodmer, founder of the Bodmer Library of World 
Literature at Cologny, a suburb of Geneva. One of the oldest 
considerable portions of the Greek New Testament is a papy- 
rus codex of the Gospel of John, the Bodmer Papyrus II, which 
was published in 1956 by Victor Martin, Professor of Classical 

' C. H. Roberts, An Unpublished Fragment of the Fourth Gospel in the John Rylands 
Library (Manchester, 1935). This was republished, with slight alterations, in the 
Bulletin of the John Rylands Library, xx (1936), pp. 45-55, and again, with critical 
notes and bibliography of reviews and opinions expressed by other scholars, in the 
Catalogue of the Greek and Latin Papyri in the John Rylands Library, iii (Manchester, 
1938), pp. 1-3. 

2 Deissmann was convinced that p" was written well within the reign of Hadrian 
(a.d. 117-38) and perhaps even during the time of Trajan (a.d. 98-117); see his 
article, 'Ein Evangelienblatt aus den Tagen Hadrians', Deutsche allgemeine £eitung, 
Nr. 564 (3 Dec. 1935); Eng. trans, in British Weekly, 12 Dec. 1935, p. 219. 

40 Important Witnesses to the 

Philology at the University of Geneva. According to its editor, 
the manuscript dates from about a.d. 200. 1 It measures about 
6 by 5^- inches and consists of six quires, of which 104 pages re- 
main. These contain the text of John i. i-vi. nandvi. 35b-xiv. 15. 
Subsequently fragments of forty-six other pages of the same 
codex were also acquired by M. Bodmer, and were edited as a 
Supplement by Martin (1958). 2 Since most of these fragments are 
small, some of them mere scraps, the amount of text of John 
xivKxxi which has been preserved is not great. 

The text of p 66 is a mixed text, with elements which are 
typically Alexandrian and Western. It is noteworthy that the 
manuscript contains about four hundred and forty alterations, 
introduced between lines, over erasures, and in the margins. 
Most of them appear to be the scribe's corrections of his own 
hasty blunders, though others seem to imply the use of a dif- 
ferent exemplar. Several passages present unique readings that 
previously had not been found in any other manuscript. In xiii. 5 
a picturesque word is used in connexion with the washing of the 
disciples' feet; according top 66 Jesus took not a 'basin' (vmTfjpa) 
but a 'foot-basin' (TroSovnrTTJpa) . In vii. 52 the presence of the 
definite article in a difficult passage now supports what scholars 
had long thought was the required sense, namely 'Search [the 
Scriptures] and you will see that the prophet does not rise from 

p 72 . What is the earliest known copy of the Epistle of Jude and 
the two Epistles of Peter is contained in another papyrus codex 
that was acquired by M. Bodmer and edited by Michel Testuz 
(1959). This manuscript, which the editor dates in the third 
century, contains a miscellaneous assortment of documents in 
the following order : the Nativity of Mary, the apocryphal cor- 
respondence of Paul to the Corinthians, the eleventh Ode of 
Solomon, the Epistle of Jude, Melito's Homily on the Pass- 
over, a fragment of a hymn, the Apology of Phileas, Psalms 

■ Herbert Hunger, the director of the papyrological collections in the National 
Library at Vienna, dates p 66 earlier, in the middle if not even in the first half of the 
second century ; see his article, 'Zur Datierung des Papyrus Bodmer II (P 66) ', Anzeiger 
der osterreichischen Akademie der Wissenschaften, phil.-hist. Kl., i960,' Nr. 4, pp. 12-33. 

2 A new edition of the Suppliment, augmented and corrected, was published in 
1962 with the assistance of J. W. B. Barns of Oxford, accompanied by a photo- 
graphic reproduction of the entire manuscript (chs. i-xxi). For still further emen- 
dations, see Barns, 'Papyrus Bodmer II, Some Corrections and Remarks', Musdon, 
lxxv (1962), pp. 327-9. See below, p. 263. 

Text of the Mew Testament 41 

xxxiii and xxxiv, and the two Epistles of Peter. Because of the 
relatively small size of the codex (it measures 6 by 5f inches), 
the editor concludes that it was made for private usage and not 
for reading in church services. Apparently four scribes took part 
in producing the manuscript. The textual affinities of its text of 
1 Peter belong definitely with the Alexandrian group, and par- 
ticularly with codex Alexandrinus. (See below, p. 263.) 

p 7 +. The Bodmer Papyrus XVII, edited by Rodolphe Kasser 
in 1 96 1, is a rather voluminous papyrus codex dating from the 
seventh century. Originally the manuscript contained 264 pages, 
each measuring about 13 by 8 inches. Today it is in a poor state 
of preservation, and contains, with many lacunae, portions of 
the Book of Acts, James, 1 and 2 Peter, 1, 2, and 3 John, and 
Jude. The type of text which it preserves agrees frequently with 
Alexandrian witnesses. (See below, pp. 263-4.) 

p 75 . Still another early Biblical manuscript acquired by M. 
Bodmer is a single-quire codex of Luke and John. It originally 
contained about 144 pages, each measuring io\ by 5^ inches, of 
which 102 have survived, either in whole or in part. The script 
is a clear and carefully executed uncial, somewhat like that of 
p 45 , though with a less pronounced ductus. The editors, Victor 
Martin and Rodolphe Kasser, date this copy between a.d. 175 
and 225. It is thus the earliest known copy of the Gospel ac- 
cording to Luke and one of the earliest of the Gospel accord- 
ing to John. The orthography of the name 'John' varies in a 
curious manner. In Luke it is invariably written with a single v 
i^Iwav-qs), and the same orthography appears at the beginning 
of the Gospel according to John. At John i. 26, however, a 
second v is added between the lines above the a and the v (as 
also at x. 40) , and thereafter the geminated form appears every- 
where except at iii. 27, where, perhaps because of a distraction, 
the scribe reverted to his former spelling. 

The textual significance of this newly acquired witness is hard 
to overestimate, presenting, as it does, a form of text very simi- 
lar to that of Vaticanus (see below, p. 264). Occasionally it is 
the only known Greek witness which agrees with the Sahidic 
in supporting several interesting readings. Thus, at John x. 7, 
instead of the traditional text, T am the door of the sheep', p 75 
replaces 'door' (77 dvpa) by 'shepherd' (6 noc^-qv). What is still 
more remarkable is the addition at Luke xvi. 19, where in 

42 Important Witnesses to the 

Jesus' account of the Rich Man and Lazarus this new witness 
inserts after ^Aou'crto? the words oVo/zan Nevr/s (see Plate III). 
The Sahidic version agrees with a rather widespread tradition 
among ancient catechists of the Coptic Church that the name 
of the Rich Man was 'Nineveh', a name which had become the 
symbol of dissolute riches. Obviously the scribe of p 75 was ac- 
quainted with this tradition and by accidental haplography 
wrote 'Neve' for 'Nineveh' {Nevrjs for Nwevrjs). 1 


X. Primacy of position in the list of New Testament manu- 
scripts is customarily given to the fourth-century codex of 
the Greek Bible which was discovered about the middle of the 
nineteenth century by Dr. Constantin von Tischendorf at the 
monastery of St. Catharine on Mount Sinai. Hence this manu- 
script is known as codex Sinaiticus. It once contained the entire 
Bible written in a carefully executed uncial hand (see Fig. i) 
and arranged with four columns per page, measuring about 
15 by 13^ inches. Today parts of the Old Testament have per- 
ished, but fortunately the entire New Testament has survived. 
In fact, codex Sinaiticus is the only known complete copy of the 
Greek New Testament in uncial script. 

The story of its discovery makes a fascinating tale, and de- 
serves to be told in some detail. In 1844, when he was not yet 

1 It was probably the horror vacui which led more than one reader to provide 
a name for the anonymous Rich Man. Toward the close of the fourth century 
Priscillian, a highly educated layman who revived certain Manichean errors in 
southern Spain, gave the name Finees to the Rich Man, perhaps because in the Old 
Testament Eleazar [compare 'Lazarus'] and Phinehas are associated. (The only 
manuscript extant of Prise. Tract, xi [ed. G. Schepps, p. 91] reads Fineet, with the 
t cancelled and s written above.) 

The widespread use of 'Dives' as the name of the Rich Man is, of course, to be 
accounted for by the rendering of the passage in the Latin Vulgate: 'Homo quidam 
erat dives et . . . .' Among the literature on the subject, see A. Harnack, 'Der Name 
des reichen Mannes in Luc 16, 19', Texte und Untersuchungen, xiii (1) (1895), pp. 75- 
78; J. Rendel Harris, 'On Certain Obscure Names in the New Testament', 
Expositor, 6th ser., i (1900), pp. 161-77, 304-8, especially pp. 175-7; Hugo 
Gressmann, Vom reichen Mann und armen Lazarus: eine literargeschichtliche Studie (Abhand- 
lungen der preussischen Akademie der Wissenschaften, phil.-hist. Kl., 19 18): A. Meyer, 
'Namen der Namenlosen', in E. Hennecke, Neutestamentliche Apokryphen, 2te Aufl. 
(Tubingen, 1 924) , pp. 78-8 1 ; L. Th. Lefort, 'Le Nom du mauvais riche (Lc 16. 19) 
et la tradition copte', ^eitschrift fiir die neutestamentliche Wissenschqft, xxxvii (1938), 
pp. 65-72; and Henry J. Cadbury, 'A Proper Name for Dives' ', Journal of Biblical 
Literature, lxxxi (1962), pp. 399-402. See also footnote 3 on p. 188 below. 

Text of the New Testament 43 

thirty years of age, Teschendorf, a Privatdozent in the Univer- 
sity of Leipzig, began an extensive journey through the Near East 
in search of Biblical manuscripts. While visiting the monas- 
tery of St. Catharine at Mount Sinai, he chanced to see some 
leaves of parchment in a waste-basket full of papers destined to 
light the oven of the monastery. On examination these proved 
to be part of a copy of the Septuagint version of the Old Testa- 
ment, written in an early Greek uncial script. He retrieved 
from the basket no fewer than forty-three such leaves, and 
the monk casually remarked that two basket loads of similarly 
discarded leaves had already been burned up! Later, when 
Tischendorf was shown other portions of the same codex (con- 
taining all of Isaiah and 1 and 4 Maccabees), he warned the 
monks that such things were too valuable to be used to stoke 
their fires. The forty-three leaves which he was permitted to 
keep contained portions of 1 Chronicles, Jeremiah, Nehemiah, 
and Esther, and upon returning to Europe he deposited them in 
the university library at Leipzig, where they still remain. In 
1846 he published their contents, naming them the codex 
Frederico-Augustanus (in honour of the King of Saxony, 
Frederick Augustus, the discoverer's sovereign and patron). 

In 1853 Tischendorf revisited the monastery of St. Catharine, 
hoping to acquire other portions of the same manuscript. The 
excitement which he had displayed on the occasion of his dis- 
covery during his first visit had made the monks cautious, and 
he could learn nothing further about the manuscript. In 1859 
his travels took him back once more to Mount Sinai, this time 
under the patronage of the Czar of Russia, Alexander II. The 
day before he was scheduled to leave he presented to the steward 
of the monastery a copy of the edition of the Septuagint which 
he had recently published in Leipzig. Thereupon the steward 
remarked that he too had a copy of the Septuagint, and pro- 
duced from a closet in his cell a manuscript wrapped in a red 
cloth. There before the astonished scholar's eyes lay the trea- 
sure which he had been longing to see. Concealing his feelings, 
Tischendorf casually asked permission to look at it further that 
evening. Permission was granted, and upon retiring to his room 
Tischendorf stayed up all night in the joy of studying the manu- 
script — for, as he declared in his diary (which as a scholar he 
kept in Latin), quippe dormire nefas videbatur ('it really seemed a 

44 Important Witnesses to the 

sacrilege to sleep') ! He soon found that the document contained 
much more than he had even hoped; for not only was most of 
the Old Testament there, but also the New Testament was in- 
tact and in excellent condition, with the addition of two early 
Christian works of the second century, the Epistle of Barnabas 
(previously known only through a very poor Latin translation) 
and a large portion of the Shepherd of Hermas, hitherto known 
only by title. 

. The next morning Tischendorf tried to buy the manuscript, 
but without success. Then he asked to be allowed to take it to 
Cairo to study ; but the monk in charge of the altar plate ob- 
jected, and so he had to leave without it. 

Later, while in Cairo, where the monks of Sinai have also 
a small monastery, Tischendorf importuned the abbot of the 
monastery of St. Catharine, who happened to be in Cairo at 
the time, to send for the document. Thereupon swift Bedouin 
messengers were sent to fetch the manuscript to Cairo, and it 
was agreed that Tischendorf would be allowed to have it quire 
by quire (i.e. eight leaves at a time) to copy it. Two Germans 
who happened to be in Cairo and who knew some Greek, an 
apothecary and a bookseller, helped him transcribe the manu- 
script, and Tischendorf revised carefully what they copied. In 
two months they transcribed 110,000 lines of text. 

The next stage of the negotiations involved what may be 
called euphemistically 'ecclesiastical diplomacy'. At that time 
the highest place of authority among the monks of Sinai was 
vacant. Tischendorf suggested that it would be to their advan- 
tage if they would make a gift to the Czar of Russia, whose 
influence, as protector of the Greek Church, they desired in 
connexion with the election of the new abbot — and what could 
be more appropriate as a gift than this ancient Greek manu- 
script! After prolonged negotiations the precious codex was 
delivered to Tischendorf for publication at Leipzig and for 
presentation to the Czar in the name of the monks. In the East 
a gift demands a return (compare Genesis xxiii, where Ephron 
'gives' Abraham a field for a burying plot, but nevertheless 
Abraham pays him 400 shekels of silver for it) . In return for 
the manuscript the Czar presented to the monastery a silver 
shrine for St. Catharine, sent a gift of 7,000 roubles for the 
library at Sinai and a gift of 2,000 roubles for the monks in Cairo, 

Text of the New Testament 45 

and conferred several Russian decorations (similar to honorary 
degrees) on the authorities of the monastery. In 1862, on the 
one-thousandth anniversary of the founding of the Russian 
Empire, the text of the manuscript was published in magnificent 
style at the expense of the Czar in four folio volumes, being 
printed at Leipzig with type cast for the purpose so as to re- 
semble the characters of the manuscript, which it represents 
line for line with the greatest attainable accuracy. 1 

The definitive publication of the codex was made in the 
twentieth century when the Oxford University Press issued a 
facsimile from photographs taken by Professor Kirsopp Lake 
(New Testament, 191 1 ; Old Testament, 1922). After the revo- 
lutions in Russia, the U.S.S.R., not being interested in the 
Bible and being in need of money, negotiated with the Trustees 
of the British Museum for the sale of the codex for £100,000 
(then slightly more than $500,000). The British Government 
guaranteed one-half the sum, while the other half was raised by 
popular subscription, contributions being made by interested 
Americans as well as individuals and congregations throughout 
Britain. Just before Christmas Day, 1933, the manuscript was 
carried under guard into the British Museum. A most thorough 
palaeographical study of the manuscript was then undertaken 
by H. J. M. Milne and T. C. Skeat of the staff of the Museum, 
and their results were published in a volume entitled Scribes 
and Correctors of Codex Sinaiticus (London, 1938). 2 Additional in- 
formation regarding the manuscript was brought to light. For 
example, the application of a new technique in manuscript 
study, the use of ultraviolet-ray lamps, enabled Milne and 
Skeat to discover that when the original scribe finished writing 
John xxi. 24 he drew two decorative lines (called a coronis) at 

1 Certain aspects of the negotiations leading to the transfer of the codex to 
the Czar's possession are open to an interpretation that reflects adversely upon 
Tischendorf 's candour and good faith with the monks at St. Catharine's. For a recent 
account intended to exculpate him of blame, see Erhard Lauch's article, 'Nichts 
gegen Tischendorf, in Bekennlnis zur Kirche: Festgabe fur Ernst Sommerlalh zum yo. 
Geburtstag (Berlin, c. 1961), pp. 15-24; for an account that includes a hitherto un- 
known receipt given by Tischendorf to the authorities at the monastery promising 
to return the manuscript from St. Petersburg 'to the Holy Confraternity of Sinai 
at its earliest request', see Ihor Sevcenko's article, 'New Documents on Tischendorf 
and the Codex Sinaiticus', published in the journal Scriptorium, xviii (1964), 
pp. 55-80. 2 See below, p. 264. 

46 Important Witnesses to the 

the lower part of the column of writing and then appended a 
subscription signifying that the text of the Gospel of John was 
finished. (Similar decorative lines and subscriptions appear else- 
where in the manuscript at the end of books.) Later the same 
scribe washed the vellum and added the concluding verse (vs. 25), 
repeating the coronis and subscription at a correspondingly 
lower position (see Plate IV). 

The type of text witnessed by Sinaiticus belongs in general 
to the Alexandrian group, but it also has a definite strain of the 
Western type of readings. Before the manuscript left the scrip- 
torium it was gone over by several scribes who did the work of 
a Siopdcorqs (corrector). Readings for which they are responsible 
for introducing are- designated in a critical apparatus by the 
siglum X\ At a later date (probably some time about the sixth 
or seventh century) a group of correctors working at Caesarea 
entered a large number of alterations into the text of both Old 
and New Testaments. These readings, designated by the siglum 
K ca or S cb , represent a thoroughgoing effort to correct the text to 
a different standard, which according to a colophon at the end of 
the books of Esdras and Esther is stated to have been 'a very 
ancient manuscript that had been corrected by the hand of the 
holy martyr Pamphilus'. 1 

A. This handsome codex, dating from about the fifth century, 
contains the Old Testament, except for several mutilations, and 
most of the New Testament (the whole of Matthew's Gospel as 
far as xxv. 6 is lost, as well as the leaves which originally con- 
tained John vi. 50-viii. 52 and 2 Cor. iv. 13-xii. 6). It was pre- 
sented in 1627 by Cyril Lucar, Patriarch of Constantinople, to 
King Charles I of England. Today it rests along with codex 
Sinaiticus in one of the prominent show-cases in the Department 
of Manuscripts of the British Museum. A photographic repro- 
duction of the codex was published in 1879-83 by the British 
Museum, under the supervision of E. Maunde Thompson. Sub- 
sequently F. G. Kenyon edited a reduced facsimile of the New 
Testament (1909) and of parts of the Old Testament. 

1 Pamphilus of Caesarea, who was martyred a.d. 309, devoted many years in 
hunting for and obtaining possession of books illustrative of the Scriptures from all 
parts of the world (Jerome, Epist. xxxiv). His library, which was especially rich in 
Biblical codices, was catalogued by Eusebius himself (Hist. Eccl. vi. 32). Among its 
priceless treasures were the Hexapla and the Tetrapla of Origen in the original 

Text of the New Testament 47 

The quality of the text preserved in codex Alexandrinus 
varies in different parts of the New Testament. In the Gospels 
it is the oldest example of the Byzantine type of text, which is 
generally regarded as an inferior form of text. In the rest of the 
New Testament (which may have been copied by the scribe 
from a different exemplar from that which he employed for the 
text of the Gospels) , it ranks along with B and X as represen- 
tative of the Alexandrian type of text. (See below, p. 264.) 

B. One of the most valuable of all the manuscripts of the 
Greek Bible is codex Vaticanus. As its name indicates, it is in 
the great Vatican Library at Rome, which has been its home 
since some date prior to 1475, when it was mentioned in the 
first catalogue made of the treasures of the library. For some 
reason which has never been fully explained, during a large part 
of the nineteenth century the authorities of the library put 
continual obstacles in the way of scholars who wished to study 
it in detail. It was not until 1889-90 that a complete photo- 
graphic facsimile of the whole manuscript, edited by Giuseppe 
Gozza-Luzi, made its contents available to all. Another fac- 
simile edition of the New Testament was issued at Milan in 
1904. (See below, p. 264.) 

The manuscript was written about the middle of the fourth 
century and contained both Testaments as well as the books of 
the Apocrypha, with the exception of the books of Maccabees. 
■ Today there are three lacunae in the codex: at the beginning 
almost forty-six chapters of Genesis are missing; a section of 
some thirty Psalms is lost; and the concluding pages (from 
Heb. ix. 14 onwards, including 1 and 2 Timothy, Titus, Phile- 
mon, and Revelation) are gone. 

The writing is in small and delicate uncials, perfectly simple 
and unadorned. Unfortunately, the beauty of the original writ- 
ing has been spoiled by a later corrector, who traced over every 
letter afresh, omitting only those letters and words which he 
believed to be incorrect. The complete absence of ornamen- 
tation from Vaticanus has generally been taken as an indi- 
cation that it is slightly older than codex Sinaiticus. On the 
other hand, some scholars believe that these two manuscripts 
were originally among the fifty copies Of the Scriptures which 
the Emperor Gonstantine commissioned Eusebius to have 
written (see pp. 7-8 above). Indeed, T. G. Skeat of the 

48 Important Witnesses to the 

British Museum has suggested to the present writer that codex 
Vaticanus was a 'reject' among the fifty copies, for it is deficient 
in the Eusebian canon tables, has many corrections by differ- 
ent scribes, and, as was mentioned above, lacks the books of 
Maccabees apparently through an oversight. Whether 'reject' 
or not, however, the text which it contains has been regarded by 
many scholars as an excellent representative of the Alexandrian 
text-type of the New Testament. 

. In common with other manuscripts of the New Testament, 
the Scripture text of Vaticanus is divided into what may be 
called chapters. The system of division, however, appears to be 
more ancient than that current in other early parchment copies 
of the New Testament. For example, in the Epistles no notice is 
taken of 2 Peter, and therefore the system of divisions appears to 
date from a time when this Epistle was not regarded as canonical. 
Furthermore, the Epistles of Paul have chapter numbers that 
do not begin afresh with each Epistle (as is customary in other 
manuscripts), but continue in one series from Romans onwards. 
In this manuscript the Epistle to the Hebrews follows theThessa- 
lonian Epistles; nevertheless, the sequence of the chapter num- 
bers discloses that in an ancestor of Vaticanus Hebrews stood 
immediately after Galatians (compare the sequence of Epistles 
hip* 6 ). 1 

C. Codex Ephraemi is the name given to a fifth-century 
manuscript of the Greek Bible that, as was mentioned in 
Chapter I, was erased during the twelfth century and many of 
its sheets rewritten with the text of a Greek translation of 
thirty-eight ascetical treatises or sermons by St. Ephraem, a 

1 In codex Vaticanus the Epistle to the Galatians concludes with the 58th 
chapter, whereas the next Epistle, that to the Ephesians, commences with the 70th 
chapter, and then the numbers continue regularly through Philippians, Colossians, 
1 and 2 Thessalonians, ending with the 93rd chapter. Following 2 Thessalonians 
stands Hebrews, which begins with the 59th chapter, and proceeds with the 60th, 
61st, 62nd, 63rd, and 64th chapter, as far as Heb. ix. 14, where the manuscript 
breaks off, the remaining part being lost. It is clear from the sequence of chapter 
divisions that in an ancestor of Vaticanus Hebrews stood after Galatians, and that 
the scribe of Vaticanus copied mechanically the chapter numerals even though they 
no longer were appropriate after Galatians. 

Since the present sequence of the New Testament books in codex Vaticanus 
agrees with the list included in Athanasius' Thirty-Ninth Festal Letter (which was 
written a.d. 367), it has sometimes been supposed that the manuscript was perhaps 
'made by Alexandrian scribes for the Emperor Gonstans during Athanasius' stay 
in Rome in 340' (Berthold Altaner [and] Alfred Stuiber, Patrologie; Leben, Schriften 
und Lehre der Kirchenvater, 7te Aufl. [Freiburg, 1966], p. 277). 

Text of the New Testament 49 

Syrian Church Father of the fourth century. By the applica- 
tion of chemical reagents and by dint of painstaking labour, 
Tischendorf was able to decipher the almost totally obliterated 
underwriting of this palimpsest. 1 Only sixty-four leaves are left 
of the Old Testament, and of the New Testament there are 
145 leaves (about five-eighths of the number which must have 
been originally required), containing portions of every book ex- 
cept 2 Thessalonians and 2 John. 

Though the document dates from the fifth century, its text 
is of less importance than one might have assumed from its 
age. It seems to be compounded from all the major text-types, 
agreeing frequently with the later Koine or Byzantine type, 
which most scholars regard as the least valuable type of New 
Testament text. Two correctors, referred to as C 2 or C b and C 3 
or C c , have made corrections in the manuscript. The former prob- 
ably lived in Palestine in the sixth century, and the latter seems 
to have done his work in Constantinople in the ninth century. 

D. Different in several respects from all the manuscripts men- 
tioned above is codex Bezae (also known as codex Cantabrigien- 
sis), which was presented in 1581 to the library at Cambridge 
University by Theodore Beza, the celebrated French scholar 
who became the successor of Calvin as leader of the Genevan 
Church. Dating from the fifth or possibly sixth century, 2 this 
codex contains most of the text of the four Gospels and Acts, 
with a small fragment of 3 John. The text is presented in Greek 
and in Latin, the two languages facing each other on opposite 
pages, the Greek being on the left, the Latin on the right. Each 
page contains a single column of text, which is not written 
straight ahead but is divided into /«2Aa, that is, lines of varying 
length with the object of making the pauses in sense come at the 
end of lines. The Gospels stand in the so-called Western order, 
with the two apostles first and the two companions of the apostles 
following (Matthew, John, Luke, and Mark). In each book the 
first three lines are in red letters, and red ink and black ink 
alternate in the lines of the subscriptions at the close of each 
book. In 1864 F. H. Scrivener published a careful edition of the 
manuscript, with full annotations, and in 1899 the Cambridge 

1 A list of corrections of Tischendorf 's edition (Leipzig, 1843), based on a fresh 
examination of the manuscript, was published by Robert W. Lyon in New Testa- 
ment Studies, v (1958-9), pp. 266-72. 2 See below, p. 264. 

5° Important Witnesses to the 

University Press issued a handsome facsimile reproduction of 
the entire manuscript. 1 

No known manuscript has so many and such remarkable 
variations from what is usually taken to be the normal New 
Testament text. Codex Bezae's special characteristic is the free 
addition (and occasional omission) of words, sentences, and 
even incidents. Thus in Luke vi this manuscript has verse 5 
after verse 10, and between verses 4 and 6 it contains the follow- 
ing account: 'On the same day, seeing one working on the 
Sabbath day, he [Jesus] said to him, "Man, if you know what 
you are doing, you are blessed; but if you do not know, you are 
accursed and a transgressor of the law" ' (see Plate V) . Although 
this sentence, which is found in no other manuscript, cannot be 
regarded as part of the original text of Luke, it may well embody 
a first-century tradition, one of the 'many other things which 
Jesus did' but which were not written in the Gospels (see John 
xxi. 25). In the Lucan account of the Last Supper (xxii. 15-20) 
this manuscript (along with some Latin and Syriac witnesses) 
omits the latter part of verse 19 and the whole of 20, thus re- 
moving all mention of the second cup, and leaving the order of 
institution inverted (Cup-Bread). In Luke xxiii. 53 there is the 
additional information that Joseph of Arimathea, after laying 
the body of Jesus in his rock-hewn tomb, 'put before the tomb a 
[great] stone which twenty men could scarcely roll'. 

Codex Bezae is the principal authority, being supported by 
one other uncial, 0, the Old Latin and Curetonian Syriac ver- 
sions, and by a few copies of the Vulgate, in inserting after 
Matt. xx. 28 the long passage: 

But seek to increase from that which is small, and to become less 
from that which is greater. When you enter into a house and are 
summoned to dine, do not sit down at the prominent places, lest per- 
chance a man more honourable than you come in afterwards, and 
he who invited you come and say to you, 'Go down lower' ; and you 
shall be ashamed. But if you sit down in the inferior place, and one 
inferior to you come in, then he that invited you will say to you, 'Go 
up higher' ; and this will be advantageous for you. 2 

1 A convenient presentation of the distinctive text of codex Bezae (collated 
against the Gebhardt-Tischendorf edition of the New Testament) was given by 
Eberhard Nestle in his Novi Testamenti supplementum (Leipzig, 1896), pp. 7-66. 

2 It may be mentioned that English translations of codex Bezae have been pub- 
lished by William Whiston (The Primitive New Testament [London, 1745]) and by 

Text of the New Testament 51 

It is particularly in the Acts of the Apostles that Bezae differs 
markedly from other witnesses, being nearly one-tenth longer 
than the text generally received. Thus in Acts xii. 10 this docu- 
ment refers to the seven steps from the prison where Peter was 
kept down to the street. In xix. 9 it adds the detail that in 
Ephesus Paul preached daily in the lecture hall of Tyranhus 
'from eleven o'clock to four', that is, at that time of day when 
the rhetorician Tyrannus would normally not hold his own 
sessions. In the Decree which the Apostolic Council at Jerusalem 
issued about a.d. 50, this manuscript omits from the list of four 
prohibitions the clause 'and from what is strangled', and adds 
at the close (Acts xv. 20 and 29) the negative Golden Rule. 

These examples will be sufficient to indicate the character- 
istic freedom of what is called the Western text, of which codex 
Bezae is the principal representative. More study has been 
expended upon this manuscript, particularly where the Greek 
text differs from the parallel Latin text and where one or both 
differ from other witnesses, than has been devoted to any other 
New Testament witness. 1 There is still no unanimity of opinion 
regarding the many problems which the manuscript raises. 

D p . The symbol D p (sometimes called D 2 ) refers to the sixth- 
century 2 codex Claromontanus, which contains only the Paul- 
ine Epistles (including Hebrews). Like codex Bezae (which 
lacks the Pauline Epistles), D p is a bilingual Greek and Latin 
manuscript, having the Greek on the left-hand page and the 
text arranged in lines of irregular length corresponding to the 
pauses in the sense. The work of at least nine different correctors 
has been identified ; the fourth of these added accent and breath- 
ing marks in the ninth century. Like codex Bezae, the type of 
text in this manuscript is distinctly Western ; it should be noted, 
however, that Western readings in the Epistles are not so strik- 
ing as those in the Gospels and Acts. An edition of the manu- 
script was published by Tischendorf in 1852. 

Johannes Greber {The New Testament; a New Translation and Explanation Based on 
the Oldest Manuscripts [New York, 1937]). J. M. Wilson published The Acts of the 
Apostles, Translated from the Codex Bezae . . . (London, 1923). See below, p. 264. 

1 See, for example, A. F.J. Klijn, A Survey of the Researches into the Western Text of 
the Gospels and Acts (Utrecht, [1949]), continued in Novum Testamentum, iii (1959), 
pp. 1-27, 161-74. Happily there is now available a complete Concordance to the Dis- 
tinctive Greek Text of Codex Bezae, compiled by James D. Yoder (Leiden and Grand 
Rapids, 1 961). 2 See below, p. 264. 

52 Important Witnesses to the 

E. Codex Basiliensis, dating from the eighth century, con- 
tains the four Gospels on 318 leaves. It is now, as its name indi- 
cates, in the library of the University of Basle, Switzerland. It 
has a Byzantine type of text. (See p. 264 and Plate VIII.) 

E a (also called E 2 ) . Formerly in the possession of Archbishop 
Laud, codex Laudianus 35 of the Bodleian Library at Oxford 
dates from the late sixth or early seventh century. It contains the 
Book of Acts in Latin and Greek, arranged in very short lines 
of only, one to three words each with the Latin in the left-hand 
column. The text exhibits a mixture of types of text, sometimes 
agreeing with codex Bezae but more often with the Byzan- 
tine type. It is the earliest known manuscript which contains Acts 
viii. 37 (the Ethiopians confession of faith; see Plate Via). An 
edition of the manuscript was published by Tischendorf in 1870. 

E p (also called E 3 ). Codex Sangermanensis, now in Lenin- 
grad, contains the Pauline Epistles in Greek and Latin on op- 
posite pages. It was copied in the ninth or tenth century from 
codex Claromontanus, and therefore is of no independent value. 

F. Codex Boreelianus, since 1830 in the University Library of 
Utrecht, once belonged to Johannes Boreel, Dutch Ambassador 
at the Court of James I of England. It contains the four Gospels 
(with large lacunae), and dates from the ninth century. Its text 
is typically Byzantine. 

Fp (also called F 2 ). Codex Augiensis, of the ninth century, 
contains the Pauline Epistles in double columns of Greek and 
Latin (Hebrews is given in Latin only). The manuscript once 
belonged to the monastery of Reichenau near Constance, which 
was known as Augia Maior; today it is in the library of Trinity 
College, Cambridge. Its text, which was published by F. H. A. 
Scrivener in 1859, is of the Western type. 

G. Codex Wolfii A, also called codex Harleianus, of the tenth 
century, contains the four Gospels with many lacunae. It was 
brought from the East by Andrew E. Seidel in the seventeenth 
century and was acquired by J. C. Wolf, who published extracts 
of it in 1723. Later it became part of the library of Robert 
Harley, and is now in the British Museum. Its text is Byzantine. 

Gp (also called G 3 ). Codex Boernerianus, once owned by the 
Leipzig professor C. F. Borner, is now at Dresden. Dating from 
the ninth century, it contains the Pauline Epistles in Greek, 
with a literalistic Latin translation between the lines. After 

Text of the New Testament 53 

Philemon there stands the superscription for the Epistle to 
the Laodiceans, but the text of this apocryphal Epistle is not 
present. In type of text it is closely akin to F p , and both of them 
probably go back one or two generations to a common arche- 
type. 1 In many respects it resembles the St. Gall MS. A, and 
is thought to have been written in the monastery of St. Gall by 
some of the Irish monks who emigrated to those parts. At the 
foot of one of the leaves (fol. 23) are eight lines of Irish verse 
which refer to making a pilgrimage to Rome : 

To come to Rome, [to come to Rome,] 

Much of trouble, little of profit, 

The thing thou seekest here, 

If thou bring not with thee, thou findest not.* 

H. Codex Wolni B was brought from the East with G, and 
passed with it into the possession of J. C. Wolf. Its subsequent 
history is unknown until 1838, when it was acquired by the 
public library at Hamburg. Dating from the ninth or tenth 
century, it contains the four Gospels with many lacunae. The 
text is of the Byzantine type. 

H a (also called H 2 ) . Codex Mutinensis, a ninth-century copy 
of Acts (lacking about seven chapters), is in the Grand Ducal 
Library at Modena. The text is of the Byzantine type. 

Hp (also called H 3 ) . Codex Coislinianus is an important codex 
of the Pauline Epistles written in a very large hand with only 
a few words in each line. The text is Alexandrian. Dating from 
the sixth century, it came into the possession of the monastery of 
the Laura on Mount Athos, where, after it began to be dilapi- 
dated, its leaves were used to supply materials for the binding 
of several other volumes. Forty-one leaves are known to exist 
today, divided among libraries at Paris, Leningrad, Moscow, 
Kiev, Turin, and Mount Athos. A note appended to the Epistle 
to Titus states that it was corrected from the copy in the library 
of Caesarea, written by the hand of the holy Pamphilus himself. 
The text is arranged according to the colometrical edition of the 
Epistles prepared by Euthalius (or Evagrius), which is found in 
several other manuscripts (see p. 26 above). 

1 See W. H. P. Hatch, 'On the Relationship of Codex Augiensis and Codex 
Boernerianus of the Pauline Epistles', Harvard Studies in Classical Philology, lx (1951), 
pp. 187-99. 

2 For the entire text, see F. H. A. Scrivener, A Plain Introduction to the Criticism of 
the New Testament, 4th ed., i (1894), p. 180, n. 2. 

828158 £ 

54 Important Witnesses to the 

I. The Washington manuscript of the Pauline Epistles, in the 
Freer Museum at the Smithsonian Institution, originally con- 
tained about 2 10 leaves, of which only 84 survive in fragmentary 
condition. Dating from the fifth or sixth century, it contains 
portions of all the Pauline Epistles except Romans (the Epistle to 
the Hebrews follows 2 Thessalonians) . The text, which was edited 
by H. A. Sanders in 1 92 1 , is a good representative of the Alexand- 
rian group, agreeing more closely with X and A than with B. 

K. Codex Cyprius, dating from the ninth or tenth century, 
is a complete copy of the four Gospels with a typically Byzantine 
type of text. 

K ap (also called K 2 ). Codex Mosquensis is a ninth- or tenth- 
century manuscript of the Acts, the Catholic Epistles, and the 
Pauline Epistles (including Hebrews). The text, which is writ- 
ten in uncial script, is separated into paragraphs by comments, 
written in minuscule script. At the foot of the page are scholia, 
attributed to John Chrysostom (see Plate IX). The text is a 
form of von Soden's /-text (see p. 141 below). 

L. Codex Regius is an eighth-century codex of the Gospels, 
nearly complete, now in the Bibliotheque Nationale at Paris. It 
was edited by Tischendorf in 1 846. Though badly written by a 
scribe who committed many ignorant blunders, its type of text 
is good, agreeing very frequently with codex Vaticanus (B). Its 
most notable feature is the presence of two endings to the Gos- 
pel according to Mark. The second of these is the traditional 
verses 9-20, but the first is a shorter ending, which is also found 
in a small number of other witnesses. This shorter ending reads 
as follows: 'But they [the women] reported briefly to Peter and 
those with him all that they had been told. And after this, Jesus 
himself sent out by means of them, from east to west, the sacred 
and imperishable proclamation of eternal salvation.' 

L ap (also called L 2 ). Codex Angelicus, now in the Angelican 
Library at Rome, is a ninth-century copy of the Acts, Catholic 
Epistles, and Pauline Epistles. Its text is mainly Byzantine. 

M. Codex Campianus, containing the four Gospels, is now 
in the Bibliotheque Nationale at Paris. It dates from the ninth 
century, and contains a text that is mainly Byzantine, but with 
admixture of Caesarean readings as well. 

N. One of the de luxe parchment manuscripts is codex 
Purpureus Petropolitanus, written in the sixth century in silver 

Text of the Mew Testament 55 

letters upon purple vellum, with gold ink for the contractions of 
the names of God and Jesus. Originally containing the four Gos- 
pels on approximately 462 leaves, it was dismembered about 
the twelfth century, possibly by Crusaders, and its leaves were 
carried far and wide. Today 182 leaves are in the Imperial 
Library at Leningrad, 33 at Patmos, 6 in the Vatican Library, 
4 in the British Museum, 2 at Vienna, 1 in private possession at 
Lerma, Italy, 1 in the Byzantine Museum, Athens, and 1 in the 
Pierpont Morgan Library, New York. 1 The text belongs pre- 
dominantly to the Byzantine type, but it preserves a number of 
readings of earlier types; B. H. Streeter regarded it (along with 
three other purple manuscripts, Z, O, and 0) as a weak mem- 
ber of the Caesarean text. 2 

O. Codex Sinopensis is a de luxe edition, written in the sixth 
century with gold ink on purple vellum, of which forty-three 
leaves of the Gospel according to Matthew survive (mainly 
chs. xiii-xxiv), as well as five miniatures. It was acquired at 
Sinope in Asia Minor by a French officer in 1899, and is now 
in the Bibliotheque Nationale at Paris. Its text, which was 
edited by Henri Omont in 1901, is a tertiary witness to the 
Caesarean type of text. 

P apr (also called P 2 ) . Codex Porphyrianus, now in Leningrad, 
a palimpsest dating from the ninth century, is one of the very few 
uncial manuscripts that include the Book of Revelation (see 
n. 3 on p. 34 above). In addition it contains the Book of Acts 
and the Catholic and Pauline Epistles, though with lacunae. The 
upper writing, which is dated in the year 1 30 1 , consists of the 
commentary of Euthalius on the Acts and the Pauline Epistles, 
together with the Biblical text. Tischendorf edited the manu- 
script in 1865-9. According to von Soden, in Acts the text is 
Koine with sporadic /-readings, and in the other books it is 
Alexandrian. According to Schmid, however, in Revelation its 
text is a secondary development of a basic Andreas type of text. 

R. Codex Nitriensis, now in the British Museum, is "a palim- 
psest containing parts of Luke in a fine large hand of the sixth 
century, over which the Syriac treatise of Severus of Antioch 

1 These last two leaves have recently been edited for the first time by Stanley 
Rypins in the Journal of Biblical Literature, Ixxv (1956), pp. 27-39. 

2 B. H. Streeter, 'Codices 157, 1071 and the Caesarean Text', Quantulacumque, 
Studies Presented to Kirsopp Lake (London, 1937), pp. 149-50. 

56 Important Witnesses to the 

against Johannes Grammaticus was written in the eighth or 
ninth century. The same manuscript also contains a palimpsest 
of 4,000 lines of Homer's Iliad. Along with more than 500 other 
manuscripts it was brought to England in 1847 from the mon- 
astery of St. Mary the Mother of God, in the Nitrian Desert, 
70 miles north-west of Cairo. According to von Soden, the text 
belongs to his / (i.e. Western) type of text. Tischendorf edited 
the text in 1857. 

• S. This is one of the earliest dated Greek manuscripts of the 
Gospels ; a colophon states that it was written by a monk named 
Michael in the year of the world 6457 (= a.d. 949). It is now in 
the Vatican Library (no. 354). The text-type is Byzantine. 

S ap . Codex Athous, of the Laura of St. Athanasius, contains 
the Acts, Catholic Epistles and Romans, and portions of 1 and 
2 Corinthians and Ephesians. Dating from the eighth or ninth 
century, its text-type is Byzantine. 

T. Codex Borgianus, at the Collegium de Propaganda Fide in 
Rome, is a valuable Graeco-Sahidic manuscript of the fifth cen- 
tury. 1 Unfortunately, it is fragmentary, preserving only 1 79 verses 
of Luke xxii-xxiii and John vi-viii. The text is very close to that 
represented by codex Vaticanus (B). 

V. Codex Mosquensis, formerly of the monastery of Vatopedi 
on Mount Athos and now in Moscow, is a copy of the four 
Gospels, nearly complete, dating from the eighth or ninth cen- 
tury. The manuscript is written in uncials down to John viii. 39, 
where it breaks off, and from that point the text is continued in 
a minuscule hand of the thirteenth century. The text-type of 
the manuscript is Byzantine. 

W. Among the more important uncial manuscripts discovered 
during the twentieth century is a codex of the four Gospels ac- 
quired by Mr. Charles L. Freer of Detroit in 1906 and now 
in the Freer Museum of the Smithsonian Institution in Washing- 
ton, D.C. It dates from the late fourth or early fifth century and, 
like codex Bezae, contains the Gospels in the so-called Western 
order (Matthew, John, Luke, and Mark). Each of the two leaves 
which serve as covers has two painted panels depicting two of 
the Evangelists ; these miniatures are thought to date from about 
the seventh or eighth century. The type of text in this manu- 

1 The text of the fragment of John was edited by A. A. Georgi, Fragmentum 
Evangelii S. Johannis Graecum Copto-Sahidicum . . . (Rome, 1789). 

Text of the New Testament 57 

script is curiously variegated, as though copied from several 
manuscripts of different families of text. In Matthew and Luke 
viii. 1 3-xxiv. 53 the text is of the common Byzantine variety, 
but in Mark i. i-v. 30 it is Western, resembling the Old Latin; 
Mark v. 31-xvi. 20 is Caesarean, akin to p 45 ; and Luke i. 1- 
viii. 12 and John v. 12-xxi. 25 are Alexandrian. The text of John 
i. i-v. 11, which fills a quire that was added about the seventh 
century, presumably to replace one which was damaged, is a 
mixed text with some Alexandrian and a few Western readings. 
In the opinion of its editor, Henry A. Sanders, this stratification 
of different kinds of text is to be explained by the theory that the 
codex goes back to an ancestor made up of fragments from 
different manuscripts of the Gospels pieced together after the 
attempt of the Emperor Diocletian to crush Christianity by 
destroying its sacred books. 

One of the most noteworthy of the variant readings in codex 
W is a remarkable insertion near the close of the Gospel accord- 
ing to Mark, part of which was known to Jerome, who declares 
that it was present 'in certain copies and especially in Greek 
codices'. Following the reference to the appearance of the risen 
Christ, who upbraided the eleven 'for their unbelief and hard- 
ness of heart, because they had not believed those who saw him 
after he had risen' (Mark xvi. 14), the text proceeds immediately 
with the following addition: 

And they excused themselves, saying, 'This age of lawlessness and 
unbelief is under Satan, who does not allow the truth and power of 
God to prevail over the unclean things of the spirits. 1 Therefore re- 
veal thy righteousness now' — thus they spoke to Christ. And Christ 
replied to them, 'The term of years for Satan's power has been ful- 
filled, but other terrible things draw near. And for those who have 
sinned I was delivered over to death, that they may return to the 
truth and sin no more; that they may inherit the spiritual and in- 
corruptible glory of righteousness which is in heaven.' 2 

X. Codex Monacensis, now in the University Library at 
Munich, contains portions of the four Gospels in the order 
Matthew, John, Luke, and Mark. It dates from the latter part 

1 Or, 'does not allow what lies under the unclean spirits to understand the truth 
and power of God'. 

1 The text of this passage, with notes and commentary, was edited by Caspar 
Keni Gregory in Das Freer-Logion (Leipzig, 1908). 

58 Important Witnesses to the 

of the ninth or from the tenth century. Except in Mark the text 
is interspersed with a patristic commentary, which is written in 
a contemporary minuscule hand. Though its text is mainly of 
the Byzantine type, it also contains occasional readings of an 
earlier type, akin to the Alexandrian text-type. 

Z. Codex Dublinensis is an interesting palimpsest in the 
library of Trinity College, Dublin. It consists of thirty-two leaves 
and preserves 295 verses of Matthew in large and broad uncials 
of the sixth (or possibly even fifth) century. The text agrees 
chiefly with that of codex Sinaiticus. The manuscript was edited 
by T. K. Abbott in 1880. 

A. Codex Sangallensis is a ninth-century Graeco-Latin manu- 
script, the Latin version being written between the lines of the 
Greek (see Plate XHIa). It contains the four Gospels complete 
with the exception of John xix. 17-35. I n Mark its text be- 
longs to the Alexandrian type, similar to that of L ; in the other 
Gospels, however, it belongs to the ordinary Koine or Byzantine 
type. The manuscript was edited by H. C. M. Rettig in 1836. 

©. Codex Koridethi is the name given to a manuscript of the 
Gospels which was discovered in the church of Sts. Kerykos and 
Julitta at Koridethi, located in the Caucasian Mountains near 
the Caspian Sea ; it is now at Tiflis, the capital of the Soviet 
Socialist Republic of Georgia. © is written in a rough, inelegant 
hand, by a scribe who clearly was not familiar with Greek. Its 
editors, Gustav Beermann and C. R. Gregory, date the manu- 
script in the ninth century. In Matthew, Luke, and John the 
text is similar to the type of text in most Byzantine manuscripts, 
but in Mark it is quite different; here it is akin to the type of 
text which Origen and Eusebius used in the third and fourth 
century at Caesarea. 

A. Codex Tischendorfianus III, now in the Bodleian Library 
at Oxford, contains the text of Luke and John in a ninth- 
century hand characterized by sloping Slavonic uncials. Its 
text is mainly Byzantine. At the close of the Gospels stands the 
so-called 'Jerusalem colophon' (see the description of MS. 157 

S. One of the most interesting palimpsest manuscripts is 
codex Zacynthius, a fragmentary codex preserving the greater 
part of Luke i. i-xi. 33. It was brought from the isle of Zante 
in 182 1 and is today in the library of the British and Foreign 

Text of the New Testament 59 

Bible Society in London. It is the earliest known New Testament 
manuscript that is provided with a marginal commentary, and 
it is the only one which has both text and commentary in uncial 
script. This commentary, which surrounds the single column of 
the Gospel text on three sides, is a catena of quotations from the 
exegetical writings of nine Church Fathers. The type of text is 
Alexandrian, akin to that of codex Vaticanus (B), and it has the 
same system of chapter divisions, which is peculiar to these two 
uncial manuscripts and to codex 579. Written in the seventh or 
eighth century, 1 it was erased in the twelfth or thirteenth cen- 
tury, and the sheets were re-used to receive the text of a Gospel 
lectionary. It was edited by S. P. Tregelles in 1861. 

77. Codex Petropolitanus is a copy of the four Gospels, almost 
complete (it lacks seventy-seven verses of Matthew and John). 
Dating from the ninth century, it contains a Byzantine type of 
text, being the head of a sub-family which is akin to, but not 
descended from, codex Alexandrinus. (See below, pp. 264-5.) 

27. Codex Rossanensis, containing Matthew and Mark, is 
written on thin vellum stained purple, in silver letters, the first 
three lines of each Gospel being in gold. Dating from the sixth 
century, it is the earliest known copy of Scripture which is 
adorned with contemporary miniatures in water-colours, seven- 
teen in number. 2 These include the raising of Lazarus, the driv- 
ing of the traders out of the temple, the ten virgins, the entry 
into Jerusalem, the foot-washing, the last supper, and Jesus 
before Pilate (see Plate VII). Its text (which was edited by 
O. von Gebhardt in 1883) is closely akin to that of N, agree- 
ing frequently with the Byzantine type of text, but with certain 
Caesarean readings as well. The manuscript belongs to the Arch- 
bishop of Rossano, at the southern end of Italy. 

0. Codex Beratinus of the sixth century is (like manuscripts 

1 W. H. P. Hatch has argued for a sixth-century date ('The Redating of Two 
Important Uncial Manuscripts of the Gospels — Codex Zacynthius and Codex 
Cyprius', Quantulacumque, pp. 333-8), but most scholars who have examined the 
manuscript prefer a later date; see J. H. Greenlee, 'The Catena of Codex Zacyn- 
thius', Biblica, xl (1959), pp. 992-1001. 

2 A study of the manuscript in its artistic aspects, with photographic repro- 
ductions of all the miniatures, was published by A. Haseloff, Codex Purpureus 
Rossanensis (Berlin and Leipzig, 1898). Another reproduction, with the plates in 
colour, was edited by A. Mufioz, // codice purpureo di Rossano (Rome, 1907). For its 
textual affinities, see William Sanday in Studia Biblica, i (Oxford, 1885), pp. 103- 

60 Important Witnesses to the 

N, O, and 2) a de luxe purple vellum manuscript written with 
silver ink. It contains only Matthew and Mark, with several 
considerable lacunae, and is in the possession of the church of 
St. George at Berat in Albania. Its text (which was edited by 
P. Batiffol in 1887) is generally of the Koine type, but it contains 
the long Western addition after Matt. xx. 28, already quoted as 
occurring in D. According to Streeter the manuscript is a ter- 
tiary witness to the Caesarean text. 

W. Codex Athous Laurae, as its name implies, is a manu- 
script in the monastery of the Laura on Mount Athos. Dating 
from about the eighth or ninth century, it contains the Gospels 
(from Mark ix onwards), Acts, the Catholic Epistles (in the 
unusual order of Peter, James, John, and Jude), the Pauline 
Epistles, and Hebrews (except one leaf of the last) . It agrees 
with L in giving the shorter ending of Mark before the longer 
one. According to Kirsopp Lake 1 its text in Mark is an early one, 
with readings both Alexandrian and Western, but chiefly akin 
to the group X, C, L, and A . The other Gospels are predomi- 
nantly Byzantine with a somewhat larger proportion of Alexan- 
drian readings than in A. 

Q. Codex Athous Dionysiou, a complete copy of the four 
Gospels (except Luke i. 15-28) in the monastery of Dionysius 
on Mount Athos, dates from the eighth or ninth century. Von 
Soden classed it as one of the three oldest of the manuscripts 
which in his opinion present the earliest variety of the Koine or 
Byzantine text. A collation made by Mary W. Winslow was 
published in 1932 by Kirsopp Lake and Silva New. 2 

046. Codex Vaticanus 2066, dating from the eighth or ninth 
century, contains the Book of Revelation between treatises of 
Basil and Gregory of Nyssa. Previously it was designated B r or 
B 2 , which gave rise to confusion with the famous codex Vatica- 
nus (B) . In text-type it is related to minuscules 6 1 and 69, with 
a form of text that differs from the early uncials as well as the later 
ecclesiastical text. 

0171. This numeral is given to two parchment fragments 
from Egypt dating from the fourth century and containing Luke 

■ Lake published the text of Mark and a collation of Luke, John, and Colossians 
in Stadia Biblica et Ecclesiastica, v (Oxford, 1903), pp. 94-131. 

1 Kirsopp Lake and Silva New, Six Collations of New Testament Manuscripts {Har- 
vard Theological Studies, xvii; Cambridge, Massachusetts, 1932), pp. 3-25- 

Text of the New Testament 61 

xxi. 45-47, 50-53, and xxii. 44-56, 61-63. According to La- 
grange it is an important witness in Egypt to the Western text. 1 
0220. This third- or fourth-century parchment leaf of Romans 
(iv. 5-v. 3 and v. 8-13) was purchased at Cairo in 1950 by 
Dr. Leland C. Wyman, Professor of Biology at Boston Univer- 
sity. The importance of 0220 lies in its agreement with codex 
Vaticanus everywhere except in v. 1, where it apparently reads 
the indicative e^o/iev. 2 


The more important minuscule manuscripts of the New 
Testament include the following. In several cases scholars have 
discovered that certain manuscripts exhibit such striking simi- 
larities of text-type as to suggest a close 'family' relationship. 

Fam. 1 . Early in the twentieth century Kirsopp Lake 3 iden- 
tified a family of witnesses that includes manuscripts 1, 118, 
131, and 209, all of which date from the twelfth to the four- 
teenth centuries. Textual analysis of the Gospel according to 
Mark indicates that the type of text preserved in these minus- 
cules often agrees with that of codex and appears to go back 
to the type of text current in Caesarea in the third and fourth 

Fam. 13. In 1868 a professor of Latin at Dublin University, 
William Hugh Ferrar, discovered that four medieval manu- 
scripts, namely 13, 69, 124, and 346, were closely related tex- 
tually.. His collations were published posthumously in 1877 by 
his friend and colleague, T. K. Abbott. It is known today that 
this group (the Ferrar group) comprises about a dozen members 
(including manuscripts 230, 543, 788, 826, 828, 983, 1689, and 
1709). They were copied between the eleventh and fifteenth 
centuries, and are descendants of an archetype which came 
either from Calabria in southern Italy or from Sicily. 4 One of the 
noteworthy features of these manuscripts is that they have the 
section about the adulterous woman (John vii. 53-viii. 11), not 

1 The text is edited in Pubblicazioni della Socuta Italiana, Papiri Greci e Latini, 
i (Florence, 1912), pp. 2-4, and ii (1913), pp. 22-25. 

2 The leaf was edited by W. H. P. Hatch in the Harvard Theological Review, xlv 
(1952), pp. 81-85. 3 In Texts and Studies, vii (2) (Cambridge, igo2). 

4 See Robert Devreesse, Les Manuscrits grecs de Vltalie me'ridionale (histoire, classe- 
ment, paUographie) (= Studi e testi, clxxxiii, Citta del Vaticano, 1955). 

62 Important Witnesses to the 

in the Fourth Gospel, but after Luke xxi. 38. Like fam. i, this 
family also has affinities with the Caesarean type of text. 1 

MS. 28. This eleventh-century copy of the four Gospels (with 
lacunae) is carelessly written but contains many noteworthy 
readings, especially in Mark, where its text is akin to the 
Caesarean type of text. It is in the Bibliotheque Nationale at 
Paris; a collation was published by the Lakes. 2 

MS. 33. Since the time of J. G. Eichhorn in the early nine- 
teenth century, MS. 33 has often been called 'the Queen of 
the cursives'. Now in the Bibliotheque Nationale at Paris, it 
is an important minuscule codex, containing the entire New 
Testament except the Book of Revelation and dating from the 
ninth or possibly tenth century. It is an excellent representa- 
tive of the Alexandrian type of text, but it shows also the in- 
fluence of the Koine or Byzantine type, particularly in Acts and 
the Pauline Epistles. 

MS. 61. This manuscript of the entire New Testament, dating 
from the late fifteenth or early sixteenth century, now at Trinity 
College, Dublin, has more importance historically than intrinsic- 
ally. It is the first Greek manuscript discovered which contains 
the passage relating to the Three Heavenly Witnesses (1 John v. 
7-8). It was on the basis of this single, late witness that Erasmus 
was compelled to insert this certainly spurious passage into the 
text of 1 John. The manuscript, which is remarkably fresh and 
clean throughout (except for the two pages containing 1 John v, 
which are soiled from repeated examination of this passage) , 
gives every appearance of having been produced expressly for 
the purpose of confuting Erasmus. (See below, p. 101.) 

MS. 69. Containing the entire New Testament, this manu- 
script was copied in the fifteenth century by a Greek named 
Emmanuel, from Constantinople, who worked for Archbishop 
Neville of York about 1 468.3 Written partly on vellum and 
partly on paper, it is an important member of fam. 13 (and was 
edited by T. K. Abbott with other members of that family). 
The manuscript is now in the Museum of Leicester, England. 

' Kirsopp and Silva Lake, Family 13 (The Ferrar Group) (Studies and Documents, 
xi, London and Philadelphia, 1941). See below, p. 265. 

2 The collation was published in their monograph, Family 13, pp. 117-54. 

3 M. R. James, Journal of Theological Studies, v (1904), pp. 445-7; xi (1910), 
pp. 291-2; and xii (191 1), pp. 465-6. See below, p. 265. 

Text of the New Testament 63 

MS. 81. Written in the year a.d. 1044, this manuscript, now in 
the British Museum, is one of the most important of all minus- 
cule manuscripts. It contains the text of Acts in a form which 
agrees frequently with the Alexandrian type of text. It was 
collated by Scrivener. 1 

MS. 157. This is a handsome twelfth-century codex of the 
Gospels, now in the Vatican Library, written for the Emperor 
John II Comnenus (11 18-43). I ts text-type resembles that of 
MS. 33, and was thought by Streeter to belong to the Caesarean 
text. 2 A colophon, which is also found in a dozen other manu- 
scripts {A, 20, 164, 215, 262, 300, 376, 428, 565, 686, 718, and 
107 1), states that it was copied and corrected 'from the ancient 
manuscripts at Jerusalem'. This colophon is repeated after each 
of the four Gospels. A collation of the manuscript was published 
by H. C. Hoskier.3 

MS. 383. This is a thirteenth-century codex of the Acts and 
Epistles (Catholic and Pauline) in the Bodleian Library at 
Oxford. It was collated by August Pott for his volume, Der 
abendlandische Text der Apostelgeschichte und die Wir-Quelle (Leipzig, 
1900), pp. 78-88, and was used by A. C. Clark in his recon- 
struction of the Western text of Acts. 

MS. 565. One of the most beautiful of all known manuscripts, 
565 is now in the public library at Leningrad. It is a de luxe copy 
of the Gospels, written in gold letters on purple vellum during 
the ninth or tenth century. 4 In Mark it is an ally of & in support 
of the Caesarean text. At the close of Mark it contains the 
so-called 'Jerusalem colophon' (see the description of codex 


MS. 579. This is a thirteenth-century copy of the Gospels 
in Paris. 5 In Matthew its text belongs to the late Byzantine 
group, but in the other Gospels it preserves an extremely good 

' F. H. Scrivener, An Exact Transcript of the Codex Augiensis . . . to which is added a 
Full Collation of Fifty Manuscripts (Cambridge, 1859). 

2 See the article cited in n. 2 on p. 55 above. 

3 Journal of Theological Studies, xiv (1913), pp. 78 ff., 242 ff., 359 ff. 

4 The text of Mark and a collation of Matthew, Luke, and John were published 
by Johannes Bclsheim in Christiania Videnskabs-Selskabs Forhandlinger, 1885, Nr. 9. 
Corrections of Belsheim's edition were included by H. S. Cronin in Texts and Studies, 
v (4) (Cambridge, 1899), pp. 106-8. 

5 Alfred Schmidtke published an edition of Mark, Luke, and John in his Die 
Evangelien eines alten Uncialcodex (BX-Text) nach einer Abschrift des dreizehnten Jahr- 
hunderts (Leipzig, 1903). 

64 Important Witnesses to the 

Alexandrian text which often agrees with B, X, and L. Like 
MS. L, it contains the double ending of Mark. 

MS. 614. A thirteenth-century codex of Acts and the Epistles 
(Pauline and Catholic) from Corfu, MS. 614 is now in the 
Ambrosian Library at Milan. It contains a large number of pre- 
Byzantine readings, many of them of the Western type of text. 1 

MS. 700. This eleventh- or twelfth-century codex of the Gos- 
pels, now in the British Museum, diverges 2,724 times from the 
Textus Receptus, and has besides 270 readings peculiar to it- 
self. 2 Along with one other Greek manuscript (no. 162) it has 
the remarkable reading in the Lucan form of the Lord's Prayer, 
'Thy Holy Spirit come upon us and cleanse us', instead of 
'Thy kingdom come' (xi. 2). This was also the text of the Lord's 
Prayer known to Marcion and Gregory of Nyssa. 3 

MS. 892. This is a ninth- or tenth-century codex of the four 
Gospels, acquired by the British Museum in 1887. 4 It contains 
many remarkable readings of an early type, belonging chiefly to 
the Alexandrian text. Von Soden observed that the scribe of 
892 preserved the divisions in pages and lines of its uncial 

MS. 1 07 1 . This twelfth-century copy of the four Gospels, now 
in the Laura on Mount Athos, contains the so-called 'Jerusalem 
colophon' referred to above in the description of codex 157. 5 
Streeter classified its text as a tertiary witness to the Caesarean 
type of text. 

MS. 1 241. MS. 1 241, containing the whole New Testament 
except the Book of Revelation, dates from the twelfth or thir- 
teenth century. 6 In the Gospels its text agrees frequently with 

1 An edition prepared by A. V. Valentine-Richards was published posthumously, 
with an introduction by J. M. Creed (Cambridge, 1934). 

2 A collation was published by H. C. Hoskier in A Full Account and Collation 
of the Greek Cursive Codex Evangelium 604 (London, 1 890) . 

3 For what can be said in support of the opinion that this variant reading stood 
in the original text of Luke, see Robert Leaney in Novum Testamentum, i (1956), 
pp. 1 03- 1 1 ; for a statement of the view that this form of the Lord's Prayer repre- 
sents a modification of the usual form for use at special services (such as ordination), 
see the present writer's discussion in the Twentieth Century Encyclopedia of Religious 
Knowledge, ii (1955), pp. 673 f. 

4 A collation was published by J. R. Harris in the Journal of Biblical Literature, ix 
(1890), pp. 31-59. 

5 A collation of Mark and of several chapters from the other Gospels was pub- 
lished by Kirsopp Lake in Studia Biblica et Ecclesiastica, v (Oxford, 1903), pp. 140-8. 

6 A collation of the Gospels was published by Kirsopp Lake in the volume men- 
tioned in n. 2, p. 60 above. 

Text of the New Testament 65 

C, L, A, W, and 33. According to Kirsopp Lake, in Matthew and 
Mark its text shows a larger infusion of Byzantine readings than 
in Luke and John. 

Fam. 1424. Codex 1424 is a ninth- or tenth-century copy of 
the entire New Testament, written by a monk named Sabas 
in the sequence of Gospels, Acts, Catholic Epistles, Revelation, 
and Pauline Epistles. All the books except Revelation are sup- 
plied with a commentary, which is written in the margins. 
Formerly in the monastery at Drama (Turkish Kosinitza) in 
Greece, it was probably taken thence to western Europe after 
the Balkan wars of 1912-13. It was bought by Dr. L. Franklin 
Gruber, president of the Chicago Lutheran Theological Semin- 
ary at Maywood, Illinois, and bequeathed at his death to the 
Seminary library. According to von Soden its text in the Gospels 
belongs to his 7*-group, which Streeter renamed fam. 1424 and 
classified as a tertiary witness to the Caesarean text. In addi- 
tion to MS. 1424, which is the oldest minuscule of the family, 
the other members of the family are M, 7, 27, 71, 115 (Matt., 
Mark) , 1 60 (Matt., Mark) ,179 (Matt., Mark) , 1 85 (Luke, John) , 
267, 349> 5!7> 659, 692 (Matt., Mark), 827 (Matt., Mark), 945, 
954, 990 (Matt., Mark), 1010, 1082 (Matt., Mark), 1188 (Luke, 
John), 1 194, 1207, 1223, !293> 139^ !4°2 (Matt., Mark), 1606, 
1675, and 2 1 91 (Matt., Mark). 

MS. 1739. Containing the Acts and the Epistles, this tenth- 
century manuscript was discovered at Mount Athos in 1879 by 
E. von der Goltz, and is usually known by his name. 1 It is of 
extreme importance because it contains a number of marginal 
notes taken from the writings of Irenaeus, Clement, Origen, 
Eusebius, and Basil. Since nothing is more recent than Basil, 
who lived from a.d. 329 to 379, it appears that the ancestor 
of this manuscript was written by a scribe toward the close of 
the fourth century. A colophon indicates that for the Pauline 
Epistles the scribe followed a manuscript which contained an 
Origenian text. It is, however, not of the Caesarean type but 
presents a relatively pure form of the Alexandrian type of text. 

MS. 2053. This is a thirteenth-century manuscript at Messina, 
containing the text of the Book of Revelation with Oecumenius' 
commentary on it. Along with codices A, C, and 2344, it is 

1 A collation made by Morton S. Enslin was published in the volume mentioned 
in n. 2, p. 60 above. 

66 Important Witnesses to the 

(according to Schmid 1 ) one of the best sources for the text of the 
Apocalypse, superior even to p 47 and X. 

MS. 2344. An eleventh-century codex now in the Biblio- 
theque Nationale at Paris, MS. 2344 contains Acts, the Catholic 
Epistles, Pauline Epistles, and Revelation (where it agrees fre- 
quently with MS. 2053), besides parts of the Old Testament. 


Manuscripts which are noteworthy because of their external 
format include the following. An uncial copy of the four Gos- 
pels, 2 no. 047, dating from the ninth or tenth century and now 
at Princeton University Library, has the writing on each page 
arranged in the form of a cross ; that is, the lines comprising the 
top third of the column and those of the bottom third are about 
one-half the length of the lines comprising the middle section of 
the column. 

Codex 1 6, a fourteenth-century copy of the four Gospels in 
Greek and Latin, formerly in the possession of Catherine de 
Medici and now in the Bibliotheque Nationale at Paris, is 
written in four colours of ink according to the contents. The 
general run of the narrative is in vermilion; the words of Jesus, 
the genealogy of Jesus, and the words of angels are in crimson ; 
the words quoted from the Old Testament, as well as those 
of the disciples, Zachariah, Elizabeth, Mary, Simeon, and John 
the Baptist, are in blue; and the words of the Pharisees, the 
centurion, Judas Iscariot, and the devil are in black. The words 
of the shepherds are also in black, but this may well have been 
an oversight. 

One of the smallest Greek manuscripts containing the four 
Gospels is MS. 461, now in the public library at Leningrad. 
There are 344 leaves, each of which measures 6f by 3| inches ; 
the single column of writing occupies an area of about 4^ by 
2 § inches. This manuscript is noteworthy also because it is 
the earliest dated Greek minuscule manuscript known to exist, 
having been copied in a.d. 835. 3 

1 Josef Schmid, Studien zur Geschichte des griechischen Apokalypse-Textes : s. Teil, 
Die alien Stamme (Munich, 1955), p. 24. 

2 Collated by William Sanday, Revue biblique, iv (1895), PP- 201-13. 

3 The earliest known Biblical manuscript which bears a date is apparently a 
palimpsest fragment of Isaiah in Syriac, written a.d. 459-60, now in the British 
Museum ; for a description of it, see E. Tisserant, 'Le plus ancien manuscrit 

Text of the Mew Testament 67 

Even more tiny must have been the vellum codex of the Book 
of Revelation of which only one leaf is extant (MS. 0169, now 
in the library of Princeton Theological Seminary; see Plate VIA). 
Discovered at Oxyrhynchus in Egypt and dating from the fourth 
century, the page measures only 3f by 2 J inches — truly a pocket 
edition! 1 

What is without doubt the largest Biblical codex is the so- 
called codex Gigas ('the giant codex'), now at Stockholm, 
which when lying open measures about 40 inches across the 
two pages and 36 inches high (see pp. 74 f. below). It is said 
that the hides of 160 asses were required for its production. 


The earliest versions of the New Testament were prepared by 
missionaries to assist in the propagation of the Christian faith 
among peoples whose native tongue was Syriac, Latin, or Coptic. 
Besides being of great value to the Biblical exegete as he traces 
the history of the interpretation of the Scriptures, these ver- 
sions are of no less importance to the textual critic in view 
of their origin in the second and third centuries. At the same 
time, however, it must be observed that there are certain limi- 
tations in the use of versions for the textual criticism of the New 
Testament. Not only were some of the translations prepared by 
persons who had an imperfect command of Greek, 3 but certain 

biblique dat6', Revue biblique, viii (191 1), pp. 85-92. For an early dated Greek 
uncial manuscript, see the description of codex S. 

1 One of the smallest Latin codices of the Gospels, measuring 5 by 4 inches and 
furnished with illuminations, is described by Francoise Henry, 'An Irish Manu- 
script in the British Museum (Add. 40618)', Journal of the Royal Society of Antiquaries 
of Ireland, lxxxvii (1957), pp. 147-66. Eight other examples of small books are dis- 
cussed by Patrick McGurk in his article 'The Irish Pocket Gospel Book', Sacris 
Erudiri, viii (1956), pp. 249-70. These manuscripts, all of which date from the 
seventh to the ninth century, are written in a very small script, with a large number 
of abbreviations. They are also distinguished by a capricious and irregular use of 
one, two, or three columns to the page for the same manuscript, and by all sorts 
of fanciful arrangements of the text on the page. See below, p. 265. 

2 For the history of scholarly research on the early versions, reference may be 
made to Arthur Voobus, Early Versions of the New Testament, Manuscript Studies 
(Stockholm, 1954), and, more briefly, the chapter on 'The Evidence of the Versions 
for the Text of the New Testament' by the present writer in New Testament Manu- 
script Studies, ed. by Parvis and Wikgren (Chicago, 1950). See below, p. 265. 

3 Cf. St. Augustine's complaint of early translators of the Bible into Latin, that 
'no sooner did anyone gain possession of a Greek manuscript, and imagine himself 
to have any facility in both languages (however slight that might be), than he made 
bold to translate it' (De doclr. Christ. 11. xi [16]). 

68 Important Witnesses to the 

features of Greek syntax and vocabulary cannot be conveyed in 
a translation. For example, Latin has no definite article ; Syriac 
cannot distinguish between the Greek aorist and perfect tenses ; 
Coptic lacks the passive voice and must use a circumlocution. 
In some cases, therefore, the testimony of these versions is ambi- 
guous. As for other questions, however, such as whether or not 
a given phrase or sentence was present in the Greek exemplar 
from which the translation was made, the evidence of the ver- 
sions is clear and valuable. 1 

The study of the early versions of the New Testament is com- 
plicated by the circumstance that various persons made vari- 
ous translations from various Greek manuscripts. Furthermore, 
copies of a translation in a certain language were sometimes 
corrected one against the other or against Greek manuscripts 
other than the ones from which the translation was originally 
made. Thus the reconstruction of a critical edition of an ancient 
version is often more complicated than the editing of the origi- 
nal Greek text itself. On the other hand, however, in tracing 
the internal history of a version the scholar has the advantage 
of divergent renderings (Ubersetzungsfarbe) to aid him. Greek 
text-types can be differentiated by variant readings alone, 
whereas in manuscripts of the versions the same Greek reading 
may be represented by different renderings. By means of such 
variant renderings as well as variant readings, the several 
stages in the evolution of a version may be traced in the manu- 
script tradition. 

The most significant of the early versions of the New Testa- 
ment are the following. 


Scholars have distinguished five different Syriac versions of 
all or part of the New Testament. They are the Old Syriac, the 
Peshitta (or common version), the Philoxenian, the Harclean, 
and the Palestinian Syriac version. 

(a) The Old Syriac version of the four Gospels is preserved to- 
day in two manuscripts, both of which have large gaps. They 
are (i) a parchment manuscript now in the British Museum, 

1 See A. F. J. Klijn, 'The Value of the Versions for the Textual Criticism of the 
New Testament', translated by H. H. Oliver in The Bible Translator, viii (1957), 
pp. 127-30. See below, p. 265. 

Text of the New Testament 69 

written in a clear and beautiful Estrangela hand (see Plate Xlla) ; 
it was edited by William Cureton in 1858 and is usually referred 
to as Syr c ; and (2) a palimpsest manuscript which Mrs. Agnes 
Smith Lewis discovered in the monastery of St. Catharine on 
Mount Sinai in 1892; it is called Syr s . Though these manu- 
scripts were copied in about the fifth and fourth centuries 
respectively, the form of text which they preserve dates from the 
close of the second or beginning of the third century. When the 
two manuscripts are compared it is seen that the Sinaitic Syriac 
represents a slightly earlier form of text than does the Cureton- 
ian, even though in some places it may have corruptions which 
the Curetonian has escaped. How far the text of the separated 
Gospels was influenced by the Gospel Harmony which Tatian 
prepared about a.d. i 70 (see pp. 89-92 below) has been much 
debated. In general the Old Syriac version is a representative of 
the Western type of text. 

The Old Syriac version of Acts and the Pauline Epistles has 
not survived in extenso; we know it only through citations 
made by Eastern Fathers. In the case of Acts, F. C. Conybeare 
reconstructed Ephraem's commentary from Armenian sources, 
of which a Latin translation is published in J. H. Ropes's The 
Text of Acts (London, 1926), pp. 373-453. Ephraem's text of 
the Pauline Epistles was reconstructed by Joseph Molitor, Der 
Paulustext des hi. Ephrdm (Monumenta biblica et ecclesiastica, vol. iv, 
Rome, 1938). 

Editions: William Cureton, Remains of a Very Antient Recension of the 
Four Gospels in Syriac... (London, 1858); F. Crawford Burkitt, 
Evangelion da-Mepharreshe: the Curetonian Version of the Four Gospels, 
with the Readings of the Sinai-Palimpsest . . ., 2 vols. (Cambridge, 1904) ; 
Agnes Smith Lewis, The Old Syriac Gospels... (London, 1910); 
Arthur Hjelt, Syrus Sinaiticus (Helsingfors, 1930) [photographic 
facsimile] . 

(b) The Peshitta version, or Syriac Vulgate, of the New Testament 
(Syr?) was prepared about the beginning of the fifth century, 
probably in order to supplant the divergent, competing Old 
Syriac translations. It contains only twenty-two books; 2 Peter, 
2 and 3 John, Jude, and Revelation were not translated. Un- 
til recently scholars thought that Rabbula, Bishop of Edessa 
(a.d. 41 1-3 1 ), was responsible for the Peshitta, but it is more 
likely that his revision marked an intermediate stage between the 

826158 F 

70 Important Witnesses to the 

Old Syriac text and the final form of the Peshitta. 1 Because the 
Peshitta was accepted as the standard version of the Scriptures 
by both Eastern and Western branches of Syrian Christendom, 
one must conclude that it had attained some degree of status 
prior to the split in the Syrian Church in a.d. 431. 

More than 350 manuscripts of the Peshitta New Testament 
are known today, several of which date from the fifth and sixth 
centuries. The text of the Peshitta has been transmitted with 
remarkable fidelity, so that very few significant variants exist 
among the witnesses. The textual complexion of the Peshitta ver- 
sion has not yet been satisfactorily investigated, but apparently 
it represents the work of several hands in various parts of the 
New Testament. In the Gospels it is closer to the Byzantine type 
of text than in Acts, where it presents many striking agreements 
with the Western text. 

Editions: P. E. Pusey and G. H. Gwilliam, Tetraevangelium sanctum 
iuxta simplicem Syrorum versionem ad fidem codicum . . . (Oxford, 1901) 
[based on forty-two manuscripts; with a critical apparatus and a 
Latin translation] ; The New Testament in Syriac (London, 1 905-20) 
[published by the British and Foreign Bible Society ; the Gospels are 
reprinted from Pusey and Gwilliam's text (without apparatus), and 
the rest of the New Testament was edited by Gwilliam and J. Gwynn] . 

(c) The Philoxenian and/or Harclean version(s). One of the most 
confused and confusing tangles of textual criticism involves the 
unravelling of the Philoxenian and/or Harclean version (s), 
usually abbreviated Syr ph and Syr h . The scanty evidence in 
several colophons found in certain Harclean manuscripts has 
been interpreted in quite different ways. On the one hand, it 
has been held that the Syriac version produced in a.d. 508 for 
Philoxenus, Bishop of Mabbug, by Polycarp his chorepiscopus 
was re-issued in 616 by Thomas of Harkel (Heraclea), Bishop 
of Mabbug, who merely added marginal notes derived from 
two or three Greek manuscripts. On the other hand, it has been 
held that the Philoxenian version was thoroughly revised by 
Thomas, who also added in the margin certain readings which 
he considered to be important but not worthy of inclusion in the 
text. In other words, according to the first view, there is but one 
version which was republished with variant readings noted in 

1 See Arthur Voobus, Studies in the History of the Gospel Text in Syriac (Louvain, 

Text of the Mew Testament 71 

the margin; according to the second, there are two separate 
versions entirely, the later one being provided with marginalia. 
It is not necessary to attempt to decide this complicated prob- 
lem here ; in any case during the sixth century, for the first time 
in the history of the Syriac-speaking Churches, the minor Catholic 
Epistles and Revelation were translated into Syriac.The Harclean 
apparatus of Acts is the second most important witness to the 
Western text, being surpassed in this respect only by codex 

Editions : Joseph White, Sacrorum Evangeliorum Versio Syriaca Philo- 
xeniana (Oxford, 1778); id., Actuum Apostolorum et Epistolarum tarn 
Catholicarum quam Paulinarum Versio Syriaca Philoxeniana (Oxford, 
1 799-1803); R. L. Bensly, The Harklean Version of the Epistle to the 
Hebrews, Chap. XI. 28-XIII. 25 (Cambridge, 1889); John Gwynn, 
The Apocalypse of St. John, in a Syriac Version Hitherto Unknown . . . 
(Dublin and London, 1897) ; id., Remnants of the Later Syriac Versions of 
the Bible . . . The Four Minor Catholic Epistles in the Original Philoxenian 
Version . . . and John VII. 52-VIII. 12 . . . (London and Oxford, 1909). 
[The text of the Apocalypse and of the minor Catholic Epistles is 
included in the British and Foreign Bible Society's edition of the 

{d) The Palestinian Syriac version. The translation into Christian 
Palestinian Syriac (i.e. Aramaic) is known chiefly from a lection- 
ary of the Gospels, preserved in three manuscripts dating from 
the eleventh and twelfth centuries. In addition fragments of 
the Gospels, in a continuous text, are extant, as well as scraps 
of Acts and of several of the Pauline Epistles. When this version 
(abbreviated SyrP al ) was made has been much disputed, but 
most scholars think that it dates from about the fifth century. 
Apparently it is based on a Greek text of the Caesarean type, 
and is quite independent of the other Syriac versions. 

Editions : Agnes Smith Lewis and Margaret Dunlop Gibson, The 
Palestinian Syriac Lectionary of the Gospels (London, 1 899) ; Agnes 
Smith Lewis, Codex Climaci rescriptus (Horae semiticae, viii, Cambridge, 
1908) [contains fragments of the Gospels, Acts, and Pauline Epistles] ; 
for other fragments, see Cyril Moss, Catalogue of Syriac Printed Books 
and Related Literature in the British Museum (London, 1962), to which 
should be added Nina V. Pigulevskaya, Katalog siriiskikh rukopisei 
Leningrada (= Palestinskii sbornik, issued by the Akademiya Nauk 
S.S.S.R., vi [69], i960, pp. 3-230). See below, p. 265. 

72 Important Witnesses to the 


When and where it was that the earliest attempts were made 
to translate the Bible into Latin has been much disputed. In 
the opinion of most scholars today the Gospels were first ren- 
dered into Latin during the last quarter of the second century 
in North Africa, where Carthage had become enamoured of 
Roman culture. Not long afterward translations were also made 
in Italy, Gaul, and elsewhere. The wooden and literalistic style 
that characterizes many of these renderings suggests that early 
copies were made in the form of interlinear renderings of the 
Greek (compare Plate Xllla). (See below, p. 265.) 

(a) The Old Latin version{s). During the third century many 
Old Latin versions circulated in North Africa and Europe, in- 
cluding distinctive versions which were current in Italy, Gaul, 
and Spain. Divergent renderings of the same verse (e.g. at Luke 
xxiv. 4-5 there are at least twenty-seven variant readings in the 
Old Latin manuscripts that have survived) bear out Jerome's 
complaint to Pope Damasus that there were almost as many ver- 
sions as manuscripts {tot enim sunt exemplaria paene quot codices). 1 

No codex of the entire Old Latin Bible is extant. The Gospels 
are represented by about thirty-two mutilated manuscripts, 
besides a number of fragments. About a dozen manuscripts of 
Acts are extant. There are four manuscripts and several frag- 
ments of the Pauline Epistles but only one complete manuscript 
and several fragments of the Apocalypse. These witnesses date 
from the fourth century to the thirteenth century, thus proving 
that the Old Latin version was still copied long after it had gone 
out of general use. The Old Latin manuscripts are designated in 
a critical apparatus by small letters of the Latin alphabet. 2 

The textual complexion of the Old Latin version(s) is typically 
Western. As a rule the form of the Old Latin current in Africa 
presents the greater divergencies from the Greek, and that cur- 
rent in Europe the smaller. 

The most important witnesses of the Old Latin versions are 
the following (grouped according to the African and European 
types of text) . 

1 See St. Jerome's preface (Novum opus) to his translation of the four Gospels. 

2 For a complete list, see Bonifatius Fischer, Vetus Latina: i, Verzeichnis der Sigel 
(Freiburg, 1 949) , or Teofilo Ayuso Marazuela, La Vetus Latina Hispana : i , Prolegdmenos 
(Madrid, 1953), pp. 224-7. 

Text of the New Testament 73 

African Old Latin manuscripts 

e. Codex Palatinus, designated by the symbol e, is a fifth- 
century manuscript containing portions of the four Gospels, 
written with silver ink on purple parchment. Though the type 
of text in e is basically African, it has been strongly europeanized. 
Augustine probably employed a Gospel text of this kind before 
a.d. 400. 

h. The symbol h is given to the fragmentary sixth-century 
manuscript known as the Fleury palimpsest, which contains 
about one-quarter of Acts, besides portions of the Catholic 
Epistles and the Book of Revelation. The manuscript contains 
many scribal errors, and the rendering into Latin is often very 
free; for example, the narrative of Paul's voyage, Acts xxviii. 1- 
13, appears to be a corrupt form of an abridgement made by 
the translator. 

k. The most important witness to the African Old Latin is 
codex Bobbiensis, to which the symbol k has been assigned. Un- 
fortunately, it is quite fragmentary, containing only about half 
of Matthew and Mark. It was copied about a.d. 400 in Africa 
and brought to the Irish monastery of Bobbio in northern Italy, 
where it was preserved for many centuries until it found a home 
in the National Library at Turin, where it is now. Its form of 
text agrees very closely with the quotations made by St. Cyprian 
of Carthage (about a.d. 250). According to E. A. Lowe 1 k shows 
palaeographical marks of having been copied from a second- 
century papyrus. It is noteworthy that k contains the inter- 
mediate ending of the Gospel according to Mark (see pp. 226- 
9 below) . 

European Old Latin manuscripts 

a. What is probably the oldest European manuscript of the 
Gospels is codex Vercellensis (known by the symbol a), kept 
in the cathedral treasure room at Vercelli in northern Italy. 
According to an old tradition it was written by the hand of 
St. Eusebius, Bishop of Vercelli, who was martyred in 370 or 
371. Next to k it is the most important Old Latin manuscript 
of the Gospels. 

b. Codex Veronensis (b), in the possession of the Chapter 

1 Reported by D. Plooij in the Bulletin of the Bezan Club, xi (1936), p. 11. 

74 Important Witnesses to the 

Library of the cathedral at Verona, Italy, is a purple parch- 
ment manuscript written in the fifth century with silver and 
occasionally gold ink. It contains the four Gospels (almost in 
their entirety) in the order of Matthew, John, Luke, and Mark. 
According to Burkitt's opinion it represents the type of text 
which Jerome used as the basis of the Vulgate. 

c. Codex Colbertinus, written in the twelfth century, probably 
in southern France, is now in the Bibliotheque Nationale at 
Paris, and contains the four Gospels in a mixed form of text. 
Clear traces of African readings persist in what is generally a 
European Old Latin text contaminated here and there by 
Jerome's Vulgate. 

d. The Latin side of the fifth- or sixth-century bilingual codex 
Bezae (D), though corrected here and there from the Greek 
side, preserves an ancient form of Old Latin text. Since d 
agrees occasionally with readings of k and of a when all other 
authorities differ, it witnesses to a text that was current no later 
than the first half of the third century, and may be earlier still. 

ff 2 . Codex Corbiensis is a mutilated copy of the four Gospels, 
of the fifth or sixth century, formerly belonging to the monastery 
of Corbey, near Amiens, and now in the Bibliotheque Nationale at 
Paris. It contains a form of text akin to that preserved in a and b. 

gig. What is undoubtedly one of the largest manuscripts in 
the world is appropriately named codex Gigas (the 'giant'). 1 
Each page measures about 20 inches wide and 36 inches high, 
and when the codex lies open it makes an impressive sight. 
Written in the early part of the thirteenth century at the Bene- 
dictine monastery of Podlazic in Bohemia, it was later ac- 
quired by the Imperial Treasury in Prague. When the Swedish 
army conquered the city in 1648, it was brought to Sweden and 
presented to the Royal Library in Stockholm the following year. 

In addition to the text of the entire Bible in Latin, the 'giant' 
manuscript contains Isidore of Seville's Etymologiae (a general 
encyclopedia in twenty books) , a Latin translation of Flavius 
Josephus' Antiquities of the Jews, Cosmas of Prague's Chronicle of 
Bohemia, as well as other works. It is sometimes called the Devil's 
Bible (Djavulsbibelri) because fol. 290 contains a huge picture of 
that potentate in garish colours, with horns, forked tongue, and 

1 For a description of the manuscript and its contents, see B. Dudik, Forschungen 
in Schwedenjur Mdhrens Geschichte (Brunn, 1852), pp. 207-35. 

76 Important Witnesses to the 

most capable Biblical scholar then living, Sophronius Eusebius 
Hieronymus, known today as St. Jerome, to undertake a re- 
vision of the Latin Bible. Within a year or so Jerome was able 
to present Damasus with the first-fruits of his work — a revision 
of the text of the four Gospels, where the variations had been 
extreme. In a covering letter he explains the principles which he 
followed : he used a relatively good Latin text as the basis for his 
revision, and compared it with some old Greek manuscripts. He 
emphasizes that he treated the current Latin text as conser- 
vatively as possible, and changed it only where the meaning was 
distorted. Though we do not have the Latin manuscripts which 
Jerome chose as the basis of his work, it appears that they be- 
longed to the European form of the Old Latin (perhaps they 
were similar to manuscript b). The Greek manuscripts appar- 
ently belonged to the Alexandrian type of text. 

When and how thoroughly Jerome revised the rest of the 
New Testament has been much debated. Several scholars (De 
Bruyne, Cavallera, B. Fischer) have argued that Jerome had 
nothing to do with the making of the Vulgate text of the rest 
of the New Testament, but that, by a curious twist of literary 
history, the work of some other translator came to be circulated 
as Jerome's work. The commonly accepted view, however, rests 
upon the natural interpretation of what Jerome says about his 
work of revision. In either case, it is apparent that the rest of 
the New Testament was revised in a much more cursory manner 
than were the Gospels. (See below, pp. 266-7.) 

It was inevitable that, in the course of the transmission of the 
text of Jerome's revision, scribes would corrupt his original work, 
sometimes by careless transcription and sometimes by deliber- 
ate conflation with copies of the Old Latin versions. In order 
to purify Jerome's text a number of recensions or editions were 
produced during the Middle Ages; notable among these were 
the successive efforts of Alcuin, Theodulf, Lanfranc, and Ste- 
phen Harding. Unfortunately, however, each of these attempts 
to restore Jerome's original version resulted eventually in still 
further textual corruption through mixture of the several types 
of Vulgate text which had come to be associated with various 
European centres of scholarship. As a result, the more than 
8,000 Vulgate manuscripts which are extant today exhibit the 
greatest amount of cross-contamination of textual types. 

Text of the New Testament 77 

The most noteworthy of these manuscripts include the follow- 
ing (they are usually denoted by capital letters or sometimes by 
the first syllable of their names) . 

A. Codex Amiatinus, in the Laurentian Library at Florence, 
is a magnificent manuscript containing the whole Bible. It was 
written by order of Ceolfrid, Abbot of Jarrow and Wearmouth, 
and sent by him as a gift to Pope Gregory in 7 16. Many scholars 
regard it as the best manuscript of the Vulgate. (See Plate lb.) 

C. Codex Cavensis, dating from the ninth century, is in the La 
Cava monastery near Salerno. It contains the whole Bible and 
is one of the chief representatives of the Spanish group of manu- 

D. Codex Dublinensis, or the Book of Armagh, is at Trinity 
College, Dublin. Dating from the eighth or ninth century, it 
contains the whole New Testament as well as the apocryphal 
Epistle of Paul to the Laodiceans. It presents the Irish type 
of Vulgate text, which is characterized by small additions 
and insertions. Here and there it shows signs of having been 
corrected from Greek manuscripts akin to the Ferrar group 
(fam. 13). (See below, p. 267.) 

F. Codex Fuldensis, now in the Landesbibliothek at Fulda, 
was written between a.d. 541 and 546 at Capua by order of 
Victor, the bishop of that see, and was corrected by him per- 
sonally. It contains the whole New Testament, together with the 
apocryphal Epistle to the Laodiceans. The Gospels are arranged 
in a single, consecutive narrative, in imitation of Tatian's Dia- 
tessaron (see pp. 89-91 below). Its text, which is very good, is 
akin to that of codex Amiatinus. 

M. Codex Mediolanensis, in the Ambrosian Library at Milan, 
is a Gospels manuscript of the early sixth century. In the judge- 
ment of Wordsworth and White it ranks with Amiatinus and 
Fuldensis as one of the best witnesses of the Vulgate. 

Y. The celebrated Lindisfarne Gospels, of about a.d. 700, 
now in the British Museum, is a beautifully executed codex, 
adorned with Celtic-Saxon illumination. It is furnished with an 
Anglo-Saxon interlinear gloss — the earliest form of the Gos- 
pels in the ancestor of English. Its text is closely akin to that of 

Z. Codex Harleianus, formerly in the Royal Library at Paris, 
was stolen from there, as it seems, by Jean Aymon in 1707 and 

78 Important Witnesses to the 

sold to Robert Harley, who deposited it in the British Museum. 
It is a beautifully written copy of the Gospels, dating from the 
sixth or seventh century. 

2. Codex Sangallensis, the oldest known manuscript of the Vul- 
gate Gospels, was written in Italy probably towards the close 
of the fifth century. 1 More than half of it survives at the monas- 
tery of St. Gall and in other libraries. Unfortunately, Wordsworth 
and White overlooked this important manuscript in preparing 
their edition of the Vulgate. 

J". What has been called one of the finest, if not the finest of 
purple manuscripts in existence is the Golden Gospels now in 
the J. Pierpont Morgan Library, New York. Written entirely in 
letters of burnished gold on purple parchment, this sumptuous 
codex contains a Vulgate Latin text with Northumbrian and 
Irish affinities. Previously thought to date from the close of the 
seventh or the beginning of the eighth century, it has lately been 
assigned to the tenth century. 2 

Editions: The decision of the Council of Trent (1546) to prepare 
an authentic edition of the Latin Scriptures was finally taken up by 
Pope Sixtus V, who authorized its publication in 1590. The Sixtine 
Vulgate was issued with a papal bull threatening the major ex- 
communication for violators of the commands that variant readings 
should not be printed in subsequent editions, and that the edition 
must not be modified. (According to Steinmiiller, 3 however, this 
bull 'today is commonly recognized as not having been properly and 
canonically promulgated'.) In 1592, after the death of Sixtus, Pope 
Clement VIII called in all the copies he could find and issued 
another authentic edition — differing from the former in some 4,900 
variants ! This latter edition remains the official Latin Bible text of 
the Roman Catholic Church to the present day. See below, p. 267. 

Since 1907 work has been going forward among Benedictine 
scholars on a revised edition of the Latin Vulgate; most of the Old 
Testament volumes have now been published. Another project in- 
volving a critical edition of the New Testament, with an apparatus, 

1 Lowe thinks that it may have been copied 'possibly during the lifetime of 
Jerome'; see E. A. Lowe, Codices Latini Antiquiores , vii (Oxford, 1956), p. 41. 

2 The earlier date was advocated by Wattenbach, de Rossi, Gregory, and its 
editor, H. G. Hoskier, The Golden Latin Gospels in the Library of J. Pierpont Morgan 
(New York, 1910); the later date was recently proposed by E. A. Lowe in 'The 
Morgan Golden Gospels: The Date and Origin of the Manuscript', Studies in Art 
and Literature for Belle da Costa Greene, ed. by Dorothy Miner (Princeton, 1954), 
pp. 266-79. 

3 John E. Steinmiiller,/! Companion to Scripture Studies, i (New York, ig4i),p. 192. 

Text of the New Testament 79 

was published at Oxford by a group of Anglican scholars. Begun by 
Bishop John Wordsworth and H. J. White, the first volume, con- 
taining the text of the four Gospels, was issued in 1899; the last vol- 
ume, containing the Book of Revelation, was completed by H. F. D. 
Sparks in 1954. For a judicious appraisal of the adequacy of the 
Oxford Vulgate (the volume containing the Gospels suffered parti- 
cularly from the editors' inexperience), see Bonifatius Fischer in £eit- 
schriftfiir die neutestamentliche Wissenschaft, xlvi (1955), pp. 178-96. 


Coptic is the latest form of the ancient Egyptian language, 
which until Christian times was written in hieroglyphs and their 
two derivatives, hieratic and demotic script. In the first centuries 
of the Christian era the language came to be written in Greek 
uncials, with the addition of seven characters taken from demotic. 

During the early Christian period the old Egyptian language 
was represented in at least half a dozen dialectal forms through- 
out Egypt, differing from one another chiefly in phonetics but 
also to some extent in vocabulary and syntax as well. In the 
southern part of the country, called Upper Egypt, the Sahidic 
dialect prevailed from Thebes to the south. Around the delta in 
the northern part of Egypt, called Lower Egypt, the Bohairic 
dialect was used along with Greek. At various settlements along 
the Nile between these two parts of the country there developed 
intermediate dialects, chiefly the Fayyumic (formerly known as 
Bashmuric), Memphitic (or Middle Egyptian), Achmimic, and 
sub-Achmimic (used south of Asyut). 

Of these dialects the Sahidic and the Bohairic are the most 
important for the study of early versions of the Bible. About the 
beginning of the third century portions of the New Testament 
were translated into Sahidic, and within the following century 
most of the books of the New Testament became available in 
that dialect. Indeed, to judge on the basis of widely divergent 
Sahidic texts, some parts of the Scriptures were translated at 
various times by independent translators. In general the Sahidic 
version agrees with the Alexandrian form of text, but in the 
Gospels and Acts it has many Western readings. Unfortunately, 
when Horner prepared his edition of this version only frag- 
mentary manuscripts were available. 1 Subsequently the Pierpont 

1 What has been thought to be the oldest Sahidic manuscript of those used by 
Horner is a papyrus codex containing portions of Deuteronomy, Jonah, and the 

80 Important Witnesses to the 

Morgan Library in New York acquired a large collection of 
Coptic manuscripts, most of which are complete. Among them 
is a Sahidic tetraevangelium (Morgan MS. 569) dating from the 
eighth or ninth century and preserving Matthew, Mark, and 
John in their entirety; Luke lacks fourteen leaves. The collec- 
tion of manuscripts is now available in photographic reproduc- 
tion in fifty-six folio volumes prepared by Henri Hyvernat. 
Mr. A. Chester Beatty also acquired three Sahidic manuscripts 
dating from about the sixth or seventh century. One of them 
(codex B) contains the Acts of the Apostles, followed by the 
Gospel according to John (see Plate XIV a) ; another (codex A) 
preserves the Pauline Epistles, followed by the Fourth Gospel. 
The third codex is fragmentary and contains the text of Psalms 
i-1 with the first chapter of Matthew. Thompson published an 
edition of the text of the Beatty Acts and Pauline Epistles, with a 
collation of the text of John against Horner's edition. 

The Bohairic version appears to be somewhat later than the 
Sahidic version. It survives in many manuscripts, almost all of 
them of a very late date (the earliest complete Gospel codex still 
extant was copied a.d. i i 74) . Recently M. Bodmer acquired an 
early papyrus codex containing most of the Gospel of John and 
the opening chapters of Genesis in Bohairic. 1 Although the first 
few folios are badly mutilated, beginning at about the middle of 
the fourth chapter of John the text is much better preserved. The 
editor, Rodolphe Kasser, is inclined to date the manuscript in 
the fourth century. It is of interest that passages which textual 
scholars have regarded as critically suspect (such as the state- 
ment about the angel's moving the water in John v. 3b-4, and 
the pericope de adultera, vii. 53-viii. 11) are not present in this 
manuscript. The Greek prototype of the Bohairic version appears 
to be closely related to the Alexandrian text-type. 

Acts of the Apostles (B.M. Or. 7594), ed. by E. A. W. Budge, Coptic Biblical Texts in 
the Dialect of Upper Egypt (London, 1912). Fol. io8 v contains a section written in 
Coptic but in a cursive Greek hand which Kenyon assigned (Introduction, p. lxiii) to 
about the middle of the fourth century. Ad. Hebbelynck, however, raised questions 
about the unity and the age of the manuscript {Muslon, xxxiv [1921], pp. 71-80), 
and H. Hyvernat thought a case could be made for dating it in the sixth century 
(quoted by Lagrange, Critique textuelle, ii [Paris, 1935], p. 324, n. 2). Sir Herbert 
Thompson, it should be added, accepted Kenyon 's fourth-century dating {The Coptic 
Version of the Acts of the Apostles . . . [Cambridge, 1932], p. xxi). 

1 The only other early manuscript of the New Testament in Bohairic (actually it 
is semi-Bohairic) is a fourth- or early fifth-century parchment fragment of Philip- 
pians, edited by Paul E. Kahle in Musion, lxiii (1950), pp. 147-57. 

Text of the Mew Testament 8 1 

Among the scattered manuscripts that preserve portions of 
the New Testament in the Fayyumic dialect, one of the earliest 
is a papyrus codex, now at the University of Michigan, which 
contains John vi. n-xv. n (with lacunae). According to its 
editor, Mrs. Elinor M. Husselman, the manuscript dates from 
the early part of the fourth century. In textual affinities it agrees 
roughly twice as often with the Sahidic version as it does with 
the Bohairic. 

The most significant representative of the sub-Achmimic 
version is a papyrus codex containing the Gospel of John. In 
the opinion of its editor, Sir Herbert Thompson, the manuscript 
dates from about a.d. 350-75. Like the Sahidic version, with 
which it is related, the version appears to be a representative 
of the Alexandrian type of text. 

Editions: [George Horner,] The Coptic Version of the New Testament 
in the Northern Dialect, otherwise called Memphitic and Bohairic . . ., 4 vols. 
(Oxford, 1898-1905) ; id., The Coptic Version of the New Testament in the 
Southern Dialect, otherwise called Sahidic and Thebaic . . ., 7 vols. (Oxford, 
191 1-24) [each with a literal English translation] ; Henri Hyvernat, 
Bybliothecae Pierpont Morgan Codices Coptici, photographice expressi . . . 
(Rome, 1922), 56 vols, in 63 [the contents are indexed in Winifred 
Kammerer, A Coptic Bibliography (Ann Arbor, 1950), pp. 33 f.] ; Her- 
bert Thompson, The Gospel of St. John According to the Earliest Coptic 
Manuscript (London, 1924) [sub-Achmimic dialect]; id., The Coptic 
Version of the Acts of the Apostles and the Pauline Epistles in the Sahidic Dia- 
to(Cambridge, 1 932) [Chester Beatty MSS.]; Rodolphe Kasser, £van- 
gile de JeanetGenese I-IV,2 (Louvain, 1958) [BodmerMS.] ; Elinor M. 
Husselman, The Gospel of John in Fayumic Coptic (P. Mich. inv. 3521) 
(Ann Arbor, 1962). For lists of all published manuscripts and frag- 
ments of the Coptic versions, see A. Vaschalde in Revue biblique, 
N.s.,xvi (1919), pp. 220-43, 513-31 ;xxix (1920), pp. 91-106, 241-58; 
xxx (1921), pp. 237-46; xxxi (1922), pp. 81-88, 234-58; Muse'on, xliii 
(1930), pp. 409-31 ;xlv (1932), pp. 1 i7-56;xlvi (1933), pp. 299-313; 
and W. C. Till, Bulletin of the John Rylands Library, xlii (igsg), 
pp. 220-40. A list of all known Coptic fragments (whether Biblical 
or not) down to the sixth century is included by Paul E. Kahle in his 
Balaizah, i (London, 1954), pp. 269-78. See below, pp. 267-8. 


Shortly after the middle of the fourth century, Ulfilas, often 
called the apostle to the Goths, translated the Bible from Greek 

82 Important Witnesses to the 

into Gothic. For this purpose he created the Gothic alphabet 
and reduced the spoken language to written form. The Gothic 
version is the earliest known literary monument in a Germanic 

The most nearly complete of the half-dozen extant Gothic 
manuscripts (all of which are fragmentary) is a de luxe copy 
dating from the fifth or sixth century and now in the University 
Library at Uppsala. It contains portions of all four Gospels, 
which stand in the so-called Western order (Matthew, John, 
Luke, and Mark). It is written on purple vellum in large let- 
ters of silver ink, whence the name which is commonly given 
to this manuscript, codex Argenteus, i.e. the 'silver codex' (see 
Plate Xlli). The initial lines of the Gospels and the first line of 
every section of text are in gold letters. All the other manuscripts 
of the Gothic New Testament, with the exception of a vellum 
leaf from a bilingual Gothic-Latin codex, are palimpsests. 

Ulfilas' translation is remarkably faithful to the original, 
frequently to the point of being literalistic. For the basis of his 
version Ulfilas used that form of Greek text which was current 
in Byzantium about a.d. 350, belonging to the early Koine type 
of text. Western readings, particularly in the Pauline Epistles, 
were subsequently introduced from Old Latin manuscripts. 

Editions: Wilhelm Streitberg, Die gotische Bibel, 2te Aufl. (Heidel- 
berg, 1 9 19; reprinted with additions, 5th ed. 1965) [Gothic and re- 
constructed Greek text on opposite pages] ; Codex Argenteus Upsaliensis, 
jussu senalus universitatis phototypice editus (Upsaliae, 1927). 


Sometimes called 'the Queen of the versions', the Armenian 
version is generally regarded as one of the most beautiful and 
accurate of all early translations of the Bible (see Plate XIV6). 
With the exception of the Latin Vulgate, more manuscripts of 
this version are extant than of any other early version; Rhodes 
has catalogued 1,244 copies of all or part of the New Testa- 
ment, and it is known that several hundred more are in libraries 
within the Soviet Union. Traditions differ regarding its origin. 
According to Bishop Koriun (died c. 450) and the historian 
Lazar of Pharb (c. 500), it was St. Mesrop (died a.d. 439), a 
soldier who became a Christian missionary, who created a new 

Text of the New Testament 83 

alphabet and, with the help of the Catholicus Sahak (Isaac the 
Great, 390-439), translated the version from the Greek text. 
On the other hand, Moses of Chorion, the nephew and disciple 
of St. Mesrop, says that Sahak made it from the Syriac text. 
Both views, with various modifications, have found defenders 
among modern scholars. There is some reason to think that the 
earliest Armenian version of the Gospels circulated in the form 
of a Harmony, distantly related to Tatian's Diatessaron. 

The earliest Armenian version appears to have undergone a 
revision prior to the eighth century. Whether the Greek text 
which served as the basis for the revision was predominantly 
Caesarean or Koine in textual type is a question which has not 
yet been satisfactorily answered. In any case, the text of Matthew 
and Mark in many Armenian manuscripts and even in Zohrab's 
printed edition appears to be strongly Caesarean in character. 

Editions : There is no satisfactory critical edition ; the edition most 
frequently used was prepared by the Mechitarist Yovhan Zohrabian 
(Venice, 1789; whole Bible 1805). The oldest known manuscript, a 
tetraevangelium copied a.d. 887, was photographically reproduced 
by G. Khalatheants (Moscow, 1899). For other manuscripts, see 
Erroll F. Rhodes, An Annotated List of Armenian New Testament Manu- 
scripts (Tokyo, 1959). 


Of all the early versions of the New Testament, probably 
the least well known among Western scholars is the Georgian 
version. The people of Caucasian Georgia, that rough, moun- 
tainous district between the Black Sea and the Caspian Sea, 
received the Gospel during the first half of the fourth century. 
The time and circumstances of the translation of the New Testa- 
ment into Georgian, an agglutinative language not known to be 
related to any other, are hidden in the mists of legend. Like the 
Armenian version, it is an important witness to the Caesarean 
type of text. (See below, p. 268.) 

Among the oldest known Gospel manuscripts are the Adysh 
manuscript of a.d. 897, the Opiza manuscript of 913, and the 
Tbet' manuscript of 995. In most apparatus critici the Adysh manu- 
script is cited as Geo 1 , and the testimony of the other two as 
Geo 2 (A and B). 

84 Important Witnesses to the 

Editions: The Old Georgian Version of the Gospel of Mark . . ., edi- 
ted with a Latin translation by Robert P. Blake (Paris, 1929); 
Matthew (Paris, 1933); John, edited by Blake and Maurice Briere 
(Paris, 1950); Luke, edited by Briere (Paris, 1955); Joseph Molitor, 
'Das Adysh-Tetraevangelium. Neu ubersetzt und mit altgeorgischen 
Paralleltexten verglichen', Oriens Christianus, xxxvii (1953), pp. 33- 
55; xxxviii (1954), pp. n-40; xxxix (1955), pp. 1-32; xl (1956), 
pp. 1-15; xli (1957), pp. 1-2 1 ; xlii (1958), pp. 1-18; xliii (1959), 
pp. 1— 1 6 ; G. Garitte, L'ancienne version georgienne des Actes des Apotres 
d'apres deux manuscrits de Sinai (Louvain, 1955). See below, p. 268. 


Scholars differ on the question of the date of the origin of the 
Ethiopic version; some argue for a date as early as the fourth 
century, while others attribute it to the sixth or seventh century. 
Opinion also differs as to whether the translators made use of a 
Greek or Syriac original. In any case, it is a curious fact that in 
the Epistles of Paul the version frequently agrees with p 45 with 
little or no other support. The version also shows evidence of 
later contamination from Coptic and Arabic texts. Thus the 
Ethiopic text eventually became a conglomerate with quite dis- 
parate elements standing side by side. The analyses which have 
been made of the earlier form of the Ethiopic version disclose 
a mixed type of text, predominantly Byzantine in complexion, 
but with occasional agreement with certain early Greek wit- 
nesses (p 46 and B) against all other witnesses. The little that is 
known of this version so far as the New Testament is concerned 
(the Old Testament has been studied more thoroughly) suggests 
that it deserves far more attention than it has received here- 
tofore. The earliest known manuscript, a codex of the four Gos- 
pels, dates from the thirteenth century ; most other manuscripts 
are of the fifteenth and succeeding centuries. 

Editions : The editio princeps of the Ethiopic New Testament was 
published by three Abyssinian monks, who issued their work in two 
volumes under the pseudonym Petrus Ethyops (Rome, 1 548-9) . This 
text was reprinted, with a Latin translation, in Brian Walton's Poly- 
glot Bible (1657). Other more recent editions, made for modern 
missionary purposes, were edited by T. Pell Piatt (London, 1826), 
re-edited by F. Pratorius (Leipzig, 1899; reprinted 1914), and F. 
da Bassano (Asmara, 1 920) . The readings of the editio princeps are 
referred to by using the siglum Eth ro . 

Text of the New Testament 85 


With the exception of St. Jerome, more is known of the life 
and work of SS. Cyril and Methodius, the apostles to the Slavs, 
than of any other translators of an ancient version of the Bible. 
Sons of a wealthy official in Salonica, they are credited with 
the creation of the Glagolitic alphabet, as well as the so-called 
Cyrillic alphabet. Soon after the middle of the ninth century 
they began the translation of the Gospels (probably in the form 
of a Greek lectionary) into Old Bulgarian, commonly called Old 
Slavonic. The version belongs basically, as one would expect, to 
the Byzantine type of text, but it also contains not a few earlier 
readings of the Western and Caesarean types. 1 

Editions: Josef Vajs, Evangelium sv. Matouse, text rekonstruovanf 
(Prague, 1935) [with the reconstructed underlying Greek text 
printed on opposite pages]; . . . Marka (Prague, 1935); . . . Lukdse 
(Prague, 1936); . . . Jam (Prague, 1936). 


Subsequent to the rise of Islam, various books of the New 
Testament were translated into Arabic from Greek, Syriac, 
Coptic (several dialects), Latin, and from combinations of these. 
During the thirteenth century two revisions were made of the 
Arabic version (or versions) current at Alexandria. As a con- 
sequence of such a tangled background the study of the Arabic 
versions is exceedingly complicated, and many problems remain 
to be solved. 

Fragments of the Nubian 2 and the Sogdian versions were 
edited at the beginning of the twentieth century, but as yet no 
thorough analysis has been made of their texts. The Anglo- 
Saxon version was translated from the Latin Vulgate. Four 
complete manuscripts and five fragmentary manuscripts of the 
Gospels are known; these date from the eleventh to the thir- 
teenth centuries. The famous Lindisfarne Gospels (see p. 77 
above) and the Rushworth Gospels, Latin manuscripts which 
were written toward the close of the seventh century, have had 
interlinear Anglo-Saxon glosses added three centuries later. 

1 For a survey of research on the Old Slavonic version, see B. M. Metzger, 
Chapters in the History of New Testament Textual Criticism (Leiden and Grand Rapids, 
■963). PP- 73-96. 2 See below, p. 268. 

826158 G 

86 Important Witnesses to the 

Two versions of the Gospels in Old Persian were published in 
the seventeenth and eighteenth centuries, but unfortunately next 
to no attention has been given them by modern textual critics, 
despite the suggestion by Kirsopp Lake that one of them shows 
traces of Caesar ean readings. 1 


Besides textual evidence derived from New Testament Greek 
manuscripts and from early versions, the textual critic has 
available the numerous scriptural quotations included in the 
commentaries, sermons, and other treatises written by early 
Church Fathers. 2 Indeed, so extensive are these citations that 
if all other sources for our knowledge of the text of the New 
Testament were destroyed, they would be sufficient alone for 
the reconstruction of practically the entire New Testament. 

The importance of patristic quotations lies in the circum- 
stance that they serve to localize and date readings and types of 
text in Greek manuscripts and versions. For example, since the 
quotations which Cyprian, Bishop of Carthage in North Africa 
about a.d. 250, includes in his letters agree almost always with 
the form of text preserved in the Old Latin manuscript k, 
scholars have correctly concluded that this fourth- or fifth- 
century manuscript is a descendant of a copy current about 250 
in North Africa. Occasionally it happens that a patristic writer 
specifically cites one or more variant readings present in manu- 
scripts existing in his day. Such information is of the utmost 
importance in providing proof of the currency of such variant 
readings at a given time and place. 3 

On the other hand, however, before the textual critic can use 
patristic evidence with confidence, he must determine whether 

1 For a survey of literature on these so-called secondary versions, see the present 
writer's chapter in M. M. Parvis and A. P. Wikgren, eds., New Testament Manuscript 
Studies (Chicago, 1950), pp. 25-68. To the literature cited there on the Anglo- 
Saxon version may be added N. R. Ker's Catalogue of Manuscripts Containing Anglo- 
Saxon (Oxford, 1957). See below, p. 268. 

2 For recent discussions of the significance of the patristic witnesses for the New 
Testament text, see R. P. Casey's essay in M. M. Parvis and A. P. Wikgren, eds., 
New Testament Manuscript Studies (Chicago, 1950), pp. 69-80; and M.-E. Boismard, 
'Critique textuelle et citations patristiques', Revue biblique, lvii (1950), pp. 388-408. 

3 For a list of two dozen such passages in the works of Origen, reference may be 
made to the present writer's contribution to Biblical and Patristic Studies in Memory 
of Robert Pierce Casey, ed. by J. N. Birdsall and R. W. Thomson (Freiburg, 1963), 
PP- 78-95- 

Text of the Mew Testament 87 

the true text of the ecclesiastical writer has been transmitted. As 
in the case of New Testament manuscripts, so also the treatises 
of the Fathers have been modified in the course of copying. The 
scribe was always tempted to assimilate scriptural quotations in 
the Fathers to the form of text which was current in the later 
manuscripts of the New Testament — a text which the scribes 
might well know by heart. 1 When the manuscripts of a Father 
differ in a given passage, it is usually safest to adopt the one 
which diverges from the later ecclesiastical text (the Textus 
Receptus or the Vulgate) . 

After the true text of the Patristic author has been recovered, 
the further question must be raised whether the writer intended 
to quote the scriptural passage verbatim or merely to paraphrase 
it. If one is assured that the Father makes a bona fide quota- 
tion and not a mere allusion, the problem remains whether he 
quoted it after consulting the passage in a manuscript or whether 
he relied on his memory. The former is more probable in the 
case of longer quotations, whereas shorter quotations were often 
made from memory. 2 Furthermore, if the Father quotes the 
same passage more than once, it often happens that he does so in 
divergent forms. Origen is notorious in this regard, for he seldom 
quotes a passage twice in precisely the same words. 3 Moreover, 

1 The requirements of memorizing portions of the Scriptures as prerequisite for 
ordination to the deaconate and the priesthood are specified in a Coptic ostracon 
edited by Crum. According to the ostracon, Samuel, Jacob, and Aaron, who 
applied to Bishop Abraham to be ordained as deacons, were required 'to master the 
Gospel according to John and learn it by heart by the end of Pentecost and to recite 
it'. Aphou, Bishop of Oxyrhynchus, is said to have required a deacon at ordination 
to know twenty-five Psalms, two Epistles of Paul, and a portion of a Gospel by 
heart ; a priest had to know, in addition, portions of Deuteronomy, Proverbs, and 
Isaiah (W. E. Crum, Coptic Ostraca from the Collections of the Egypt Exploration Fund 
[London, 1902], p. 9, no. 29). According to the Rules of St. Pachomius, applicants 
for entrance into the monastery were required to know twenty Psalms or two Epistles 
of Paul, or a portion of some other part of Scripture {Regulae Monasticae S. Pachomii, 
ed. by P. B. Albers, p. 41 ; see also Richard Reitzenstein, Historia Monachorum, 
pp. 61 f., 162 ff.). 

2 According to Preuschen, 'the briefer the citation is, the greater is the prob- 
ability that it represents the text favoured by Origen; the more extensive the 
citation is, so much the more doubtful is it whether we can recognize in it the text 
of Origen' (Erwin Preuschen, 'Bibelcitate bei Origenes', ^eitschrifi fur die neutesta- 
mentliche Wissenschaft, iv [1903], pp. 71 f). 

3 The memory can play strange tricks when one quotes even themost familiar 
passages. 'Dr. Salmon adduces from E. Abbot (Authorship of the Fourth Gospel, p. 39) a 
remarkable instance of this in no less a person than Jeremy Taylor, who quotes the 
text "Except a man be born again he cannot see the kingdom of God" nine times, 

88 Important Witnesses to the 

while dictating to one of his several amanuenses, Origen would 
sometimes refer merely to a few catchwords in the Scripture 
passage which had come to his mind as an illustration of his 
argument ; later the amanuensis would hunt out the passage in 
a Biblical manuscript and insert its words at the appropriate 
place in Origen's treatise. 1 Differences among the longer quo- 
tations could therefore arise, depending upon which of several 
available copies of the Scriptures the amanuensis may have con- 
sulted. Despite these difficulties which attend the determination 
and evaluation of patristic evidence for the New Testament text, 
this kind of evidence is of such great importance in tracing the 
history of the transmission of the text that the labour of refining 
the ore from the dross is well worth the effort. 

The following is a list of several of the more important Church 
Fathers whose writings contain numerous quotations from the 
New Testament. 2 

Ambrose of Milan, d. 397 

Ambrosiaster [= Pseudo- Ambrose] of Rome, second half of fourth 

Athanasius, Bishop of Alexandria, d. 373 
Augustine, Bishop of Hippo, d. 430 
Chrysostom, Bishop of Constantinople, d. 407 
Clement of Alexandria, d. c. 212 
Cyprian, Bishop of Carthage, d. 258 
Cyril of Alexandria, d. 444 
Didymus of Alexandria, d. c. 398 
Ephraem the Syrian, d. 373 
Epiphanius, Bishop of Salamis, d. 403 
Eusebius, Bishop of Caesarea, d. 339 or 340 
Gregory of Nazianzus in Cappadocia, d. 389 or 390 
Gregory of Nyssa in Cappadocia, d. 394 
Hilary of Poitiers, d. 367 
Hippolytus of Rome, d. 235 
Irenaeus, Bishop of Lyons, d. c. 202 

yet only twice in the same form, and never once correctly' (F. G. Kenyon, Hand- 
book to the Textual Criticism of the New Testament, 2nd ed., 1912, p. 245). 

1 For several such passages in Origen's commentary on John, see Preuschen in 
Die griechischen christlichen Schriftsteller , Origenes, iv, pp. lxxxix ff. 

1 For information concerning the literary contributions of these Fathers, as well 
as bibliographical references to editions and monographs, see Berthold Altaner, 
Patrology (New York, i960), and the more extensive work by Johannes Quasten, 
Patrology (Westminster, Md., 1950 ff.), of which three of the four projected volumes 
have already appeared. 

Text of the Mew Testament 89 

Isidore of Pelusium, d. c. 435 

Jerome [= Hieronymus], d. 419 or 420 

Justin Martyr, d. c. 165 

Lucifer of Galaris (Cagliari), d. 370 or 371 

Marcion, flourished at Rome, c. 150-60 

Origen of Alexandria and Gaesarea, d. 253 or 254 

Pelagius, fourth-fifth century 

Primasius, Bishop of Hadrumentum, d. soon after 552 

Pseudo-Hieronymus, fifth-sixth century 

Rufinus of Aquileia, d. 410 

Tatian, flourished c. 170 

Tertullian of Carthage, d. after 220 

Theodore of Mopsuestia in Cilicia, d. 428 

Among these writers one of the more controversial figures 
was Tatian, a Syrian from Mesopotamia, known chiefly for his 
Diatessaron, or Harmony of the four Gospels. Combining dis- 
tinctive phrases preserved by only one Evangelist with those 
preserved by another, he arranged the several sections of the 
Gospels into a single narrative. Omitting only a very few sec- 
tions (such as the genealogies of Jesus in Matthew and in Luke, 
the former of which traces Jesus' lineage from Abraham on- 
wards, and the latter of which traces it backwards to Adam), 
Tatian managed to preserve practically the entire contents of 
the separate Gospels woven into one. The name by which it 
came to be known is 'Diatessaron', 1 derived from the Greek 
phrase 8ia reaadpwv, meaning 'through [the] four [Gospels]'. 

Tatian' s Harmony soon became popular, particularly in the 
East. As late as the fifth century Theodoret, who became Bishop 
of Cyrrhus or Cyrus on the Euphrates in upper Syria in a.d. 423, 
found that many copies of the Diatessaron were in use within his 
diocese. But because Tatian had become heretical in his later 
life, and because Theodoret believed that orthodox Christians 
were in danger of being corrupted by using Tatian'swork, he 
destroyed all of the copies of the Diatessaron that he could find 
(totalling about 200) and put in their place the separate Gospels 
of the four Evangelists. 2 

x Victor of Capua's reference (in codex Fuldensis; see p. 77 above) to Tatian's 
work as a Diapente is not a lapsus calami, as some have thought, but is to be explained 
as the application to the Harmony of a musical term that involves four intervals; 
see Franco Bolgiani, Vittore di Capua e il 'Diatessaron' {Memorie dell'Accademia delle 
Scienze di Torino, Classe di scienze morali, storiche e filologiche, Serie 4 a , n. 2; 
1962). 2 Theodoret, Treatise on Heresies, i. 20. 

go Important Witnesses to the 

As a result of the zeal of Bishop Theodoret, and doubtless of 
others like him, no complete copy of Tatian's Diatessaron is ex- 
tant today. In 1933 a small fragment of parchment containing a 
portion of the Diatessaron in Greek was unearthed by archaeo- 
logists on the site of the ancient Roman fortress-town of Dura- 
Europos on the lower Euphrates. It is known that this town fell 
to the Persians under King Shapur I in a.d. 256-7, and there- 
fore the fragment must antedate that event. 1 The text which is 
preserved contains the narrative of the coming of Joseph of 
Arimathea for the body of Jesus. A literal translation will show 
how words and phrases from all four Gospels have been woven 
together. Since the left-hand margin of the vellum has suffered 
damage, the first half-dozen or so letters at the beginning of each 
line are lacking. They can be restored, however, with almost 
perfect confidence. In the following rendering the restorations 
are enclosed within square brackets and the modern Scripture 
references (which are not, of course, in the fragment) are en- 
closed within parentheses. 

[. . . the mother of the sons of Zebed]ee (Matt, xxvii. 56) and 
Salome (Mark xv. 40) and the wives [of those who] had followed 
him from [Galileje to see the crucified (Luke xxiii. 4gb-c). And [the 
da]y was Preparation; the Sabbath was dawfning] (Luke xxiii. 54). 
And when it was evening (Matt, xxvii. 57), on the Preparation], that 
is, the day before the Sabbath (Mark xv. 42), [there came] up a man 
(Matt, xxvii. 57), be[ing] a member of the council (Luke xxiii. 50), 
from Arimathea (Matt, xxvii. 57), a c[i]ty of [Jude]a (Luke xxiii. 51), 
by name Jo[seph] (Matt, xxvii. 57), good and ri[ghteous] (Luke 
xxiii. 50), being a disciple of Jesus, but se[cret]ly, for fear of the 
[Jew]s (John xix. 38). And he (Matt, xxvii. 57) was looking for [the] 
k[ingdom] of God (Luke xxiii. 51b). This man [had] not [con] sen ted 
to [their] p[urpose] (Luke xxiii. 51a). .. . 

It is evident that Tatian went about composing his Dia- 
tessaron with great diligence. Probably he worked from four 
separate manuscripts, one of each of the Gospels, and, as he wove 
together phrases, now from this Gospel and now that, he would 
no doubt cross out those phrases in the manuscripts from which 
he was copying. Otherwise it is difficult to understand how he 

1 The fragment was edited by Carl H. Kraeling in Studies and Documents, vol. iii 
(London, 1935). See below, p. 269. 

Text of the New Testament 91 

was able to put together so successfully a cento of very short 
phrases from four separate documents. 

The most spectacular reading preserved in this fragment is 
near the beginning. Although it rests partly on a restoration, 
and although none of the secondary translations of Tatian which 
were known hitherto exhibits the reading, it is probable that 
Tatian referred to 'the wives of those who had followed' Jesus 
from Galilee. This statement and the implications which it con- 
veys are without parallel in the text of the separate Gospels in 
any other manuscript or version. 

In addition to the Greek fragment, portions of Tatian's Har- 
mony are known through the quotations from it which certain 
early Syrian Church Fathers included in their homilies and other 
treatises, particularly in the commentary that St. Ephraem 
of the fourth century wrote on the Diatessaron. Until recently 
this commentary was available only in an Armenian transla- 
tion, preserved in two manuscripts that were copied a.d. i 195. 
In 1957 announcement was made of the acquisition by Sir 
Chester Beatty of a Syriac manuscript containing about three- 
fifths of Ephraem's treatise. 1 The manuscript, which dates from 
the late fifth or early sixth century, has been edited by Dom 
Louis Leloir, 2 who previously prepared a careful edition and 
translation of the commentary in its Armenian form. Several 
harmonies in other languages (Arabic and Persian in the East, 
and Latin, Medieval Dutch, Old English, and Old Italian in the 
West) show more or less dependence, either in form or in text, 
upon Tatian's pioneering work. 

How much contamination the Diatessaron has exerted on 
the transmission of the text of the separate Gospels has been 
variously estimated. Not many scholars today agree with the 
extreme views of Hermann von Soden and Anton Baumstark, 
who thought that they detected the presence of Tatianic in- 
fluence on a great number of Eastern and Western manuscripts 

1 For a description and a preliminary discussion of the manuscript, see Dom Louis 
Leloir, 'L'Originale syriaque du commentaire de S. £phrem sur le Diatessaron', 
Biblica, xl (1959), pp. 959-70 (= Studia Biblica et Orientalia, ii, pp. 391-402). Tjitze 
Baarda has edited a brief quotation of Ephraem's Commentary which is included 
in a Nestorian manuscript brought to Rome in 1869 ('A Syriac Fragment of Mar 
Ephraem's Commentary on the Diatessaron', New Testament Studies, viii [1962], 
pp. 287-300). 

2 Saint Lphrem, Commentaire de VEvangile concordant, texte syriaque (Manuscrit Chester 
Beatty yog), (Dublin, 1963). See below, p. 269. 

92 Important Witnesses to the Text 

of the Gospels. It is doubtless true, however, that not a few 
instances of harmonization of the text of the Gospels in certain 
witnesses (notably the Western witnesses) are to be ascribed to 
Tatian's influence. 1 

1 For a recent summary of the history of research on Tatian's Diatessaron, 
reference may be made to B. M. Metzger's Chapters in the History of New Testament 
Textual Criticism (Leiden and Grand Rapids, 1963), pp. 97-120. See below, p. 269. 


The History of 

New Testament Textual Criticism 

as reflected in Printed Editions 

of the Greek Testament 


The Pre-critical Period: 

The Origin and Dominance of the 

Textus Receptus 


Johannes Gutenberg's invention of printing by using 
movable type had the most momentous consequences for 
Western culture and civilization. Now copies of books could 
be reproduced more rapidly, more cheaply, and with a higher 
degree of accuracy than had ever been possible previously. 
Quite appropriately, the first major product of Gutenberg's press 
was a magnificent edition of the Bible. The text was Jerome's 
Latin Vulgate, and the volume was published at Mayence 
(Mainz) between 1450 and 1456. During the next fifty years at 
least one hundred editions of the Latin Bible were issued by 
various printing houses. In 1488 the first edition of the complete 
Hebrew Old Testament came from the Soncino press in Lom- 
bardy. Before 1500 Bibles had been printed in several of the 
principal vernacular languages of Western Europe — Bohemian 
(Czech), French, German, and Italian. 

But, except for several short extracts, 1 the Greek New Testa- 
ment had to wait until 15 14 to come from the press. Why was 
there so long a delay? Two reasons may be suggested which 
help to account for the lapse of some sixty years from Guten- 
berg's invention to the first printed Greek Testament. 

In the first place, the production of fonts of Greek type neces- 
sary for a book of any considerable size was both difficult and 
expensive. 2 The attempt was made to reproduce in print the 

1 In 1 48 1 the Greek text of the hymns of Zachariah and Elizabeth (Luke i) was 
printed at Milan in an appendix to a Greek Psalter. In 1504 the first six chapters of 
John in Greek (taken from a Gospel lectionary, and retaining the rubrics) were 
printed at Venice in a book containing a Latin translation of the Poems of Gregory 
Nazianzen. In 1514 John i. 1-14 in Greek was reprinted at Tubingen. 

2 The first book with continuous Greek type was a grammar of the Greek 
language entitled Erotemata, written by Constantine Lascaris and published at 

96 The Pre-critical Period 

appearance of minuscule Greek handwriting, with its numerous 
alternative forms of the same letter, as well as its many com- 
binations of two or more letters (ligatures). 1 Instead, therefore, 
of producing type for merely the twenty-four letters of the Greek 
alphabet, printers prepared about 200 different characters. 
(Subsequently these variant forms of the same letters were 
abandoned, until today there remain only the two forms of the 
lower-case sigma, a and ?.) 

The principal cause which retarded the publication of the 
Greek text of the New Testament was doubtless the prestige 
of Jerome's Latin Vulgate. Translations into the vernacular 
languages were not derogatory to the supremacy of the Latin 
text from which they were derived. But the publication of the 
Greek New Testament offered to any scholar acquainted with 
both languages a tool with which to criticize and correct the 
official Latin Bible of the Church. 

At length, however, in 15 14 the first printed Greek New 
Testament came from the press, as part of a Polyglot Bible. 
Planned in 1 502 by the cardinal primate of Spain, Francisco 
Ximenes de Cisneros (1437— 151 7), this magnificent edition of 
the Hebrew, Aramaic, Greek, and Latin texts was printed at 
the university town of Alcala (called Complutum in Latin) . 
Known as the Complutensian Polyglot, 2 the project was under 
the editorial care of several scholars, of whom Diego Lopez de 
Zuniga (Stunica) is perhaps best known. 3 Volume v, containing 

Milan in 1476; see Richard P. Breaden, 'The First Book Printed in Greek', Bulletin 
of the New York Public Library, li (1947), pp. 586-92. Before 1476 several editions of 
Latin classics and Church Fathers had appeared containing occasional Greek words 
and phrases. The honour of being the first to print'Greek goes either to Johann Fust 
and Peter Schoeffer for their edition of Cicero's De officii! and Paradoxa, published 
at Mainz in 1465, or to Conrad Sweynheym'and Arnoldus Pannartz for their volume 
of Lactantius' Opera, issued at Subiaco, Italy, on 30 Oct. 1465. See Victor Schol- 
derer, Greek Printing Types, 1465-1927 (London, 1927), p. 1. See below, p. 269. 

1 For a list of many of the ligatures used in incunabula, see the Style Manual of the 
United States Government Printing Office (Washington, 1945), pp. 316-18, or 
Georg F. von Ostermann, Manual of Foreign Languages (New York, 1952), pp. 105-8. 

2 Of the original 600 sets which were printed, the locations of ninety-seven are 
known today; see James P. R. Lyell, Cardinal Ximenes (London, 1917), and Mariano 
Revilla Rico, La Poliglota de Alcala, estudio historico-critico (Madrid, 1 917). 

3 One of the less well-known collaborators in the project, who nevertheless 
played an important role in the preparation of the Greek text of the Septuagint and 
of the New Testament for the Polyglot, was Demetrius Ducas, a Greek from Crete 
whom Ximenes brought to Alcala to teach in his Academy and to edit Greek books ; 
see the account concerning Ducas in Deno J. Geanakoplos, Greek Scholars in Venice: 

The Origin and Dominance of the Textus Receptus 97 

the New Testament and a Greek glossary with Latin equivalents, 
was printed first, its colophon bearing the date 10 January 1514. 
Volume vi, the appendix, containing a Hebrew lexicon and an 
elementary Hebrew grammar, was printed next, in 1515. The 
four volumes of the Old Testament appeared last, the colo- 
phon of volume iv bearing the date 10 July 151 7. The sanction 
of Pope Leo X, printed in volume i, was not obtained until 
22 March 1520, but it appears that for some reason the Poly- 
glot was not actually circulated (that is, published) until about 

The four volumes which contain the Old Testament presenV 
the Hebrew text, the Latin Vulgate, and the Greek Septuagint 
in three columns side by side on each page, together with the 
Aramaic Targum of Onkelos for the Pentateuch at the foot of 
the page, accompanied by a Latin translation. The Greek type 
used in the New Testament volume is modelled after the style of 
the handwriting in manuscripts of about the eleventh or twelfth 
century, and is very bold and elegant (see Plate XVI). 1 It 
is printed without rough or smooth breathing marks and is 
accented according to a system never heard of before or since : 
monosyllables have no accent, while the tone syllable in other 
words is marked with a simple apex, resembling the Greek acute 
accent mark. Each word or group of Greek words is coded to 
the adjacent column of the Latin Vulgate by small supralinear 
roman letters, thus assisting readers with little Greek to find the 
equivalent words in each column. The Septuagint is printed 
with the familiar cursive style of Greek characters popularized 
by Aldus Manutius, the famous Venetian printer. 

What Greek manuscripts lie behind the text of the Com- 
plutensian New Testament has never been satisfactorily 

Studies in the Dissemination of Greek Learning from Byzantium to Western Europe (Cam- 
bridge, Massachusetts, 1962), pp. 238 ff. 

1 According to Scholderer, the unknown designer of the Greek font of type used 
in the Complutensian New Testament incorporated certain features of Nicolas 
Jenson's Greek font, which was put in use in 1471. 'In fact, it [the Greek of the 
Complutensian New Testament] is the last and most beautiful example of the Jen- 
sonian class of type, carrying it to the limit of its possibilities, and is fairly entitled 
to its generally acknowledged position as the king of all Greek founts, although its 
full perfection was not disclosed until the present century, when Robert Proctor 
completed a revival of the lower-case in an enlarged copy by adding a full set of 
capitals adapted to it. The Greek of the Complutensian New Testament remains 
the only original contribution of Spain to Hellenic typography' (Scholderer, 
op. cit., p. 10). 

98 The Pre-critical Period 

ascertained. 1 In his dedication to Pope Leo X, after mentioning 
the pains which he had taken to secure Latin, Greek, and Hebrew 
manuscripts, Ximenes continues: 'For Greek copies indeed we 
are indebted to your Holiness, who sent us most kindly from the 
Apostolic Library very ancient codices, both of the Old and the 
New Testament; which have aided us very much in this under- 
taking.' 2 

Though the Complutensian text was the first Greek New 
Testament to be printed, the first Greek New Testament to be 
published (that is, put on the market) was the edition prepared 
by the famous Dutch scholar and humanist Desiderius Erasmus 
of Rotterdam (1469- 1536). It cannot be determined exactly 
when Erasmus first decided to prepare an edition of the Greek 
Testament, but on a visit to Basle in August 1 514 he discussed 
(probably not for the first time) the possibility of such a volume 
with the well-known publisher Johann Froben. Their negotia- 
tions seem to have been broken off for a time, but were resumed 
in April 15 15 while Erasmus was on a visit at the University of 
Cambridge. It was then that Froben importuned him through 
a mutual friend, Beatus Rhenanus, to undertake immediately 
an edition of the New Testament. Doubtless Froben had heard 
of the forthcoming Spanish Polyglot Bible and, sensing that 
the market was ready for an edition of the Greek New Testa- 
ment, wished to capitalize upon that demand before Ximenes' 
work would be finished and authorized for publication. Froben's 
proposal, which was accompanied by a promise to pay Erasmus 

1 See, for example, Franz Delitzsch, Studien zur Entstehungsgeschichte der Polyglotten- 
bibel des Cardinals Ximenes (Programm, Leipzig, 1 87 1 ; also published [in German] at 
London under the title, Studies on the Complutensian Polyglott, 1872) ; Samuel Berger, 
La Bible au seizieme siecle (Paris, 1879), pp. 51-58; Delitzsch, Fortgesetzte Studien zur 
Entslehungsgeschichte der Complutensischen Polyglottenbibel (Programm, Leipzig, 1 886) ; 
and M. Goguel, 'Le Texte et les Editions du Nouveau Testament grec', Revue de 
I'histoire des religions, lxxxii (1920), pp. 14-18. 

1 Some have doubted the truth of this statement, for Leo had been elected Pope 
less than a year before the New Testament volume was finished and therefore (so it 
is argued) could hardly have sent manuscripts in time to be useful (see Marvin R. 
Vincent, A History of the Textual Criticism of the New Testament [New York, 1899], 
p. 50). One should observe, however, that Ximenes does not explicitly say that the 
manuscripts were sent during Leo's pontificate; it is altogether possible, as Hug 
suggested long ago, that they had been sent during the pontificate of the previous 
Pope, Julius II, through the intervention of the Cardinal de Medici, who had great 
influence over Julius and who in turn succeeded him to the Papal throne (J. L. 
Hug, Einleitung in die Schriften des Neuen Testaments [Stuttgart und Tubingen, 1826], 
§ 55)- 

The Origin and Dominance of the Textus Receptus 99 

as much as anyone else might offer for such a job (se daturum 
pollicetur, quantum alius quisquam), apparently came at an oppor- 
tune moment. Going to Basle again in July of 1 5 1 5, Erasmus 
hoped to find Greek manuscripts sufficiently good to be sent to 
the printer as copy to be set up in type along with his own Latin 
translation, on which he had been working intermittently for 
several years. To his vexation the only manuscripts available on 
the spur of the moment required a certain amount of correcting 
before they could be used as printer's copy. (See below, p. 269.) 
The printing began on 2 October 15 15, and in a remark- 
ably short time (1 March 151 6) the entire edition was finished, 
a large folio volume of about 1,000 pages which, as Erasmus 
himself declared later, was 'precipitated rather than edited' 
(praecipitatum verius quam editum). Owing to the haste in produc- 
tion, the volume contains hundreds of typographical errors ; in 
fact, Scrivener once declared, '[It] is in that respect the most 
faulty book I know.' 1 Since Erasmus could not find a manuscript 
which contained the entire Greek Testament, he utilized several 
for various parts of the New Testament. For most of the text 
he relied on two rather inferior manuscripts from a monastic 
library at Basle, one of the Gospels (see Plate XV) and one 
of the Acts and Epistles, both dating from about the twelfth 
century. 2 Erasmus compared them with two or three others 
of the same books and entered occasional corrections for the 
printer in the margins or between the lines of the Greek script. 3 
For the Book of Revelation he had but one manuscript, dat- 
ing from the twelfth century, which he had borrowed from 
his friend Reuchlin. Unfortunately, this manuscript lacked the 
final leaf, which had contained the last six verses of the book. 
For these verses, as well as a few other passages throughout 
the book where the Greek text of the Apocalypse and the ad- 
joining Greek commentary with which the manuscript was 
supplied are so mixed up as to be almost indistinguishable, 

1 F. H. A. Scrivener, A Plain Introduction to the Criticism of the New Testament, 
4th ed., ii (London, 1894), p. 185. 

2 Though some have dated the manuscript of the Gospels in the fifteenth cen- 
tury (so Scrivener [doubtfully], Kenyon, and von Dobschiitz), Gregory, Eberhard 
Nestle, von Soden, and Clark assign it to the twelfth. 

3 These corrections are described by Kenneth W. Clark, 'The Erasmian Notes in 
Codex 2', in Studia Evangelica, ed. by K. Aland, F. L. Cross, et al. (= Texte und Unter- 
suchungen,\xxu\; Berlin, 1959), pp. 749-56. See also C. C. Tarelli, 'Erasmus's Manu- 
scripts of the Gospels', Journal of Theological Studies, xliv (1943), pp. 155-62. 

ioo The Pre-critical Period 

Erasmus depended upon the Latin Vulgate, translating this text 
into Greek. As would be expected from such a procedure, here 
and there in Erasmus' self-made Greek text are readings which 
have never been found in any known Greek manuscript — but 
which are still perpetuated today in printings of the so-called 
Textus Receptus of the Greek New Testament. 1 

Even in other parts of the New Testament Erasmus occasion- 
ally introduced into his Greek text material taken from the 
Latin Vulgate. Thus in Acts ix. 6, the question which Paul asks 
at the time of his conversion on the Damascus road, 'And he 
trembling and astonished said, Lord, what wilt thou have me 
to do?', was frankly interpolated by Erasmus from the Latin 
Vulgate. This addition, which is found in no Greek manuscript 
at this passage (though it appears in the parallel account of Acts 
xxii. 10), became part of the Textus Receptus, from which the 
King James version was made in 1611. 

The reception accorded Erasmus' edition, the first published 
Greek New Testament, was mixed. On the one hand, it found 
many purchasers throughout Europe. Within three years a 
second edition was called for, and the total number of copies 
of the 1 5 16 and 15 19 editions amounted to 3,300. The second 
edition became the basis of Luther's German translation. 2 On 
the other hand, in certain circles Erasmus' work was received 
with suspicion and even outright hostility. His elegant Latin 
translation, differing in many respects from the wording of 
Jerome's Vulgate, was regarded as a presumptuous innovation. 
Particularly objectionable were the brief annotations in which 
Erasmus sought to justify his translation. He included among 
the philological notes not a few caustic comments aimed at the 
corrupt lives of many of the priests. In the words of J. A. Froude, 
'The clergy's skins were tender from long impunity. They 

1 For example aKaOdprriTos (Rev. xvii. 4; there is, however, no such word in the 
Greek language as dKaBdprrjs, meaning 'uncleanness') ; opBptvos (xxii. 16); eXBe 
twice, iXdtrin (xxii. 17)* ou^iapTupou^tat yap . . . e-niTidfj irpos rayra (xxii. 18); 
d<f>at.pfj fiifiXov . . . d<f>aiprjoti (future for d^eAef!!), |3ij3Aou (second occurrence) 
(xxii. 19); iiLtov (xxii. 21). 

2 It has often been debated how far Luther's translation rests on the Greek 
text. In recent times H. Dibbelt (Archiv fur Reformationsgeschichte, xxxviii [1941], 
pp. 300-30) has maintained that the translation reflects only an occasional consul- 
tation of the Greek; on the other hand, H. Bornkamm (Theologische Literaturzeitung, 
lxxii [1947], pp. 23-28) holds that Luther translated from the combination of Greek 
and Latin texts in Erasmus' edition and from the Vulgate, which he had in his 
head. See below, p. 269. 

The Origin and Dominance of the Textus Receptus 101 

shrieked from pulpit and platform, and made Europe ring with 
their clamour.' 1 As a result, 'universities, Cambridge and Ox- 
ford among them, forbade students to read Erasmus's writings 
or booksellers to sell them'. 2 

Among the criticisms levelled at Erasmus one of the most 
serious appeared to be the charge of Stunica, one of the editors 
of Ximenes' Complutensian Polyglot, that his text lacked part 
of the final chapter of i John, namely the Trinitarian statement 
concerning 'the Father, the Word, and the Holy Ghost: and 
these three are one. And there are three that bear witness in 
earth' ( I John v. 7-8, King James version). Erasmus replied that 
he had not found any Greek manuscript containing these words, 
though he had in the meanwhile examined several others be- 
sides those on which he relied when first preparing his text. 
In an unguarded moment Erasmus promised that he would 
insert the Comma Johanneum, as it is called, in future editions if 
a single Greek manuscript could be found that contained the 
passage. At length such a copy was found — or was made to 
order ! As it now appears, the Greek manuscript had probably 
been written in Oxford about 1520 by a Franciscan friar named 
Froy (or Roy) , who took the disputed words from the Latin Vul- 
gate. 3 Erasmus stood by his promise and inserted the passage 
in his third edition (1522), but he indicates in a lengthy footnote 
his suspicions that the manuscript had been prepared expressly 
in order to confute him.* 

Among the thousands of Greek manuscripts of the New Tes- 
tament examined since the time of Erasmus, only three others 
are known to contain this spurious passage. They are Greg. 88, 
a twelfth-century manuscript which has the Comma written in 
the margin in a seventeenth-century hand; Tisch. cu 1 10, which 
is a sixteenth-century manuscript copy of the Complutensian 
Polyglot Greek text ; and Greg. 629, dating from the fourteenth 

1 J. A. Froude, Lift and Letters of Erasmus (New York, 1896), p. my. See also 
August Bludau, Die beiden ersten Erasmus- Ausgaben des Neuen Testaments und ihre Gegner 
(Freiburg i. Br., 1902). 

2 Froude, op. cit., p. 138. 

3 See J. Rendel Harris, The Origin of the Leicester Codex of the New Testament 
(London, 1887), pp. 40-53; and C. H. Turner, The Early Printed Editions of the 
Greek Testament (Oxford, 1924), pp. 23-24. 

4 To'day the codex (designated Greg. 61), which is in the library of Trinity 
College, Dublin, opens almost of its own accord at 1 John v — so often has it been 
consulted at this passage ! 

826158 H 

102 The Pre-critical Period 

or, as Riggenbach has argued, from the latter half of the six- 
teenth century. 1 The oldest known citation of the Comma is in 
a fourth-century Latin treatise entitled Liber apologeticus (ch. 4), 
attributed either to Priscillian or to his follower, Bishop Instan- 
tius of Spain. The Comma probably originated as a piece of 
allegorical exegesis of the three witnesses and may have been 
written as a marginal gloss in a Latin manuscript of 1 John, 
whence it was taken into the text of the Old Latin Bible during 
the fifth century. The passage does not appear in manuscripts 
of the Latin Vulgate before about a.d. 800. In view of its in- 
clusion in the Clementine edition of the Latin Vulgate (1592), 
in 1897 the Holy Office in Rome, a high ecclesiastical congre- 
gation, made an authoritative pronouncement, approved and 
confirmed by Pope Leo XIII, that it is not safe to deny that this 
verse is an authentic part of St. John's Epistle. 2 Modern Roman 
Catholic scholars, however, recognize that the words do not be- 
long in the Greek Testament; for example, the four bilingual 
editions of the New Testament that were edited by Bover, Merk, 
Nolli, and Vogels include the words as part of the Vulgate text 
approved by the Council of Trent, but reject them from the 
Greek text that faces the Latin on the opposite page. 3 

Subsequently Erasmus issued a fourth and definitive edition 
(1527), which contains the text of the New Testament in three 
parallel columns, the Greek, the Latin Vulgate, and Erasmus' 
own Latin version. He had seen Ximenes' Polyglot Bible shortly 
after the publication of his own third edition in 1522, and 
wisely decided to avail himself of its generally superior text in 
the improvement of his own. In the Book of Revelation, for 
example, he altered his fourth edition in about ninety passages 
on the basis of the Complutensian text. A fifth edition, which 
appeared in 1535, discarded the Latin Vulgate but differs very 
little from the fourth as regards the Greek text. 

Thus the text of Erasmus' Greek New Testament rests upon 
a half-dozen minuscule manuscripts. The oldest and best of 
these manuscripts (codex 1, a minuscule of the tenth century, 
which agrees often with the earlier uncial text) he used least, 
because he was afraid of its supposedly erratic text ! Erasmus' 

1 Das Comma Johanneum (Giitersloh, 1928). 2 See below, p. 269. 

3 For a full discussion by a noted Roman Catholic textual scholar, see Teofilo 
Ayuso Marazuela, 'Nuevo estudio sobre el "Comma Ioanneum" ', Biblica, xxviii 
(1947), pp. 83-112, 216-35; xx ' x ( 1 948) ) PP- 52-76. See below, p. 270. 

The Origin and Dominance of the Textus Receptus 103 

text is inferior in critical value to the Complutensian, yet because 
it was the first on the market and was available in a cheaper and 
more convenient form, it attained a much wider circulation and 
exercised a far greater influence than its rival, which had been 
in preparation from 1502 to 1514. In addition to Erasmus' five 
editions mentioned above, more than thirty unauthorized re- 
prints are said to have appeared at Venice, Strasbourg, Basle, 
Paris, and other places. 

Subsequent editors, though making a number of alterations in 
Erasmus' text, essentially reproduced this debased form of the 
Greek Testament. Having secured an undeserved pre-eminence, 
what came to be called the Textus Receptus of the New Testa- 
ment resisted for 400 years all scholarly efforts to displace it in 
favour of an earlier and more accurate text. The highlights of 
this history are as follows. 

The first edition of the whole Bible in Greek was published in 
three parts in February 1 5 1 8 at Venice by the celebrated Aldine 
press. The New Testament, which is dedicated to Erasmus, fol- 
lows the first edition of Erasmus so closely as to reproduce many 
typographical errors — even those which Erasmus had corrected 
in the list of errata ! 

A beautifully printed pocket-sized edition (its pages measure 
3 by 4 inches) was produced in two volumes (616 pp., 475 pp.) 
by Ioannes Antonius de Nicolinis de Sabio at Venice in 1538. 
Its text, edited by Melchiorre Sessa, is curiously eclectic, depend- 
ing now on Erasmus, now on the Aldine text, and occasionally 
departing from all previous editions. 1 Like several other early 
editions it contains in Greek certain 'helps for readers', such as 
lists of chapter headings, lives of the Evangelists, hypotheses 
(introductions) to the several books, and accounts of the jour- 
neys of Paul and of his martyrdom. 2 

The famous Parisian printer and publisher, Robert Estienne, 

1 For an analysis of the textual affinities of this extremely rare edition, see 
W. H. P. Hatch, 'An Early Edition of the New Testament in Greek', Harvard 
Theological Review, xxxiv (1941), pp. 69-78. The rarity of copies of the edition may 
be gauged from the statement of Reuss that after diligent search he was unable to 
locate a copy anywhere in the libraries of Germany. Hatch indicates that only seven 
copies of the complete edition, and one copy of volume ii, are known to exist today. 
(In addition to the copies mentioned by Hatch, the present writer subsequently 
acquired a copy of volume ii.) 

2 These 'helps for readers' were drawn from minuscule manuscripts of the New 
Testament; see pp. 25 f. above. 

104 The Pre-critical Period 

latinized as Stephanus (1503-59), issued four editions of the 
Greek Testament, three at Paris (1546, 1549, and 1550) and the 
last at Geneva (1551), where he spent his final years as a pro- 
fessed Protestant. 1 The three Parisian editions are most sump- 
tuously printed, with type cast at the expense of the French 
Government. The handsome third edition, of folio size (8| by 
13 inches), is the first Greek Testament that has a critical ap- 
paratus ; Stephanus entered on the inner margins of the pages 
variant readings from fourteen Greek codices as well as many 
readings from the Complutensian Polyglot. One of the manu- 
scripts that he cited is the famous codex Bezae, which had been 
collated for him, he says, 'by friends in Italy'. 

The text of Stephanus' editions of 1 546 and 1 549 was a com- 
pound of the Complutensian and Erasmian editions; the third 
edition (1550) approaches more closely the text of Erasmus' 
fourth and fifth editions. As it happened, Stephanus' third edition 
became for many persons, especially in England, the received or 
standard text of the Greek Testament. 

Stephanus' fourth edition (1551), which contains two Latin 
versions (the Vulgate and that of Erasmus) printed on either 
side of the Greek text, is noteworthy because in it for the first 
time the text was divided into numbered verses. It has often 
been stated that Stephanus marked the verse divisions while 
journeying 'on horseback', and that some of the infelicitous 
divisions arose from the jogging of the horse that bumped his 
pen into the wrong places. 2 Stephanus' son does indeed assert 
that his father did the work while on a journey {inter equitandum) 
from Paris to Lyons, but the most natural inference is that 
the task was accomplished while resting at the inns along the 
road. (See below, p. 270.) 

In 1553 Stephanus' folio edition of 1550 was reprinted in 
a small-sized volume (3I by 5^ inches) by Jean Crispin (or 
Crespin) , the French printer of Geneva, who published many 
editions of the Scriptures in various languages, including the 
second quarto English Geneva Bible of 1570. Crispin repro- 
duced the text of Stephanus with only half a dozen minor 
alterations. The variant readings of the 1550 folio edition are 

1 See Elizabeth Armstrong, Robert Estienne, Royal Printer: an Historical Study of the 
Elder Stephanus (Cambridge, 1954), pp. 211 ff. 

2 So, for example, A. T. Robertson, An Introduction to the Textual Criticism of the 
New Testament, 2nd ed. (New York, 1928), p. 100. 

The Origin and Dominance of the Textus Receptus 105 

also reproduced, 1 though without Stephanus' sigla referring to 
individual manuscripts. It was either Stephanus' folio edition 
or Crispin's pocket-sized reprint that William Whittingham 
and his fellow Protestant refugees from England utilized when 
they prepared their English translation of the New Testament 
(Geneva, 1557), the first English version to include variant read- 
ings in the margins. 2 

Theodore de Beze (Beza, 15 19-1605), the friend and succes- 
sor of Calvin at Geneva and an eminent classical and Biblical 
scholar, published no fewer than nine editions of the Greek Tes- 
tament between 1565 and 1604, and a tenth edition appeared 
posthumously in 161 1. Only four of them, however, are in- 
dependent editions (those of 1565, 1582, 1588-9, and 1598), the 
others being smaller-sized reprints. Accompanied by annotations 
and his own Latin version, as well as Jerome's Latin Vulgate, 
these editions contain a certain amount of textual information 
drawn from several Greek manuscripts which Beza had collated 
himself, as well as the Greek manuscripts collated by Henry 
Stephanus, son of Robert Stephanus. Noteworthy among Beza's 
own manuscript possessions were codex Bezae and codex 
Claromontanus, though he made relatively little use of them, 
for they deviated too far from the generally received text of the 
time. Beza seems also to have been the first scholar to collate 
the Syriac New Testament, which was published in 1569 by 
Emmanuel Tremellius. For Acts and 1 and 2 Corinthians he 
utilized information from the Arabic version put at his disposal 
by Franciscus Junius. Despite the variety of this additional 
textual evidence available to Beza, which is reflected chiefly in 
his annotations, the Greek text which he printed differs little 
from Stephanus' fourth edition of 1551. The importance of 
Beza's work lies in the extent to which his editions tended to 
popularize and to stereotype the Textus Receptus. The King 
James translators of 1 6 1 1 made large use of Beza's editions of 
1588-9 and 1598. 
In 1624 tne brothers Bonaventure and Abraham Elzevir, 

1 The variant reading at Luke xvii. 35, however, is introduced into Crispin's 
text itself. 

2 The translation, with minor alterations, and the variant readings were incor- 
porated in the Geneva Bible of 1 560 ; see the discussion of the variant readings in 
B. M. Metzger, 'The Influence of Codex Bezae upon the Geneva Bible of 1560', 
New Testament Studies, viii (1961), pp. 72-77. 

106 The Pre-critical Period 

two enterprising printers at Leiden, 1 published a small and con- 
venient edition of the Greek Testament, the text of which was 
taken mainly from Beza's smaller 1 565 edition. The preface to the 
second edition, which appeared in 1633, makes the boast that 
'[the reader has] the text which is now received by all, in which 
we give nothing changed or corrupted'. 2 Thus from what was 
a more or less casual phrase advertising the edition (what mod- 
ern publishers might call a 'blurb'), there arose the designation 
'Textus Receptus', or commonly received, standard text. Partly 
because of this catchword the form of the Greek text incor- 
porated in the editions that Stephanus, Beza, and the Elzevirs 
had published succeeded in establishing itself as 'the only true 
text' of the New Testament, and was slavishly reprinted in hun- 
dreds of subsequent editions. It lies at the basis of the King 
James version and of all the principal Protestant translations in 
the languages of Europe prior to 1881. So superstitious has been 
the reverence accorded the Textus Receptus that in some cases 
attempts to criticize or emend it have been regarded as akin to 
sacrilege. Yet its textual basis is essentially a handful of late and 
haphazardly collected minuscule manuscripts, and in a dozen 
passages its reading is supported by no known Greek witness. 


The next stage in the history of New Testament textual criti- 
cism is characterized by assiduous efforts to assemble variant 
readings from Greek manuscripts, versions, and Fathers. For 
almost two centuries scholars ransacked libraries and museums, 
in Europe as well as the Near East, for witnesses to the text of 
the New Testament. But almost all of the editors of the New 
Testament during this period were content to reprint the time- 
honoured but corrupt Textus Receptus, relegating the evidence 
for the earlier readings to the apparatus. An occasional brave 
soul who ventured to print a different form of Greek text was 
either condemned or ignored. 

The first systematic collection of variant readings (those given 
in the margin of Stephanus' 1 550 edition had been assembled 

1 The printing house of Elzevir (properly Elzevier), founded by Louis Elzevier 
(1540-1617), issued many beautiful editions of classical authors from 1595 to 1681. 

2 'Textum ergo habes, nunc ab omnibus receptum : in quo nihil immutatum aut 
corruptum damus.' 

The Origin and Dominance of the Textus Receptus 107 

somewhat at random) was included in the Polyglot Bible edi- 
ted by Brian Walton (1600-61) and published at London in 
1655-7 m s i x f° no volumes. The fifth volume (1657) contains 
the New Testament in Greek, Latin (both the Vulgate and the 
version of Arius Montanus), Syriac, Ethiopic, Arabic, and (for 
the Gospels) Persian. The Greek text as well as each of the 
oriental versions is supplied with a literal translation into 
Latin. The Greek text is that of Stephanus' 1550 edition, with 
slight alterations. At the foot of the page are variant readings 
from codex Alexandrinus, which had recently been presented 
(1627) by Cyril Lucar, the Patriarch of Constantinople, to 
Charles I. In the sixth volume of the Polyglot, the Appendix, 
Walton included a critical apparatus, prepared by Archbishop 
Ussher, of variant readings derived from fifteen other authori- 
ties, to which were added the variants from Stephanus' margin. 1 

In 1675 Dr. John Fell (1625-86), 2 Dean of Christ Church and 
afterwards Bishop of Oxford, issued anonymously a small-sized 
volume (3! by 6^ inches), the first Greek Testament to be pub- 
lished at Oxford. The text, drawn from the Elzevir 1633 edi- 
tion, was supplied with an apparatus in which Fell claimed 
to give variants from more than 100 manuscripts and ancient 
versions. Unfortunately, however, about twenty of these wit- 
nesses, including codex Vaticanus (B), are not cited indivi- 
dually, but only in statements concerning the total number 
of manuscripts which agree in any particular reading. For the 
first time evidence from the Gothic and Bohairic versions, sup- 
plied by T. Marshall, was also made available through Fell's 

About the time of the publication of Fell's edition, John Mill 3 
(164.5-1707), a fellow of Queen's College, Oxford, began his 
studies of New Testament textual criticism which were to come 
to fruition thirty years later in an epoch-making edition of the 

1 The London Polyglot was attacked by Dr. John Owen, the Puritan Dean of 
Christ Church, Oxford, in Considerations on the Prolegomena and Appendix to the late 
Polyglotta (1659), to which Walton made a prompt and sharp rejoinder, The Con- 
siderator Considered (1659). Walton's stature as a scholar and churchman was 
recognized at the Restoration, when he was appointed Bishop of Chester in 1660. 
In 1667 the London Polyglot had the distinction of being put on the Index 
Librorum Prohibitorum. 

2 This Dr. Fell is the theme of Thomas Brown's well-known quatrain (adapted 
from Martial, Epig. i. 32) beginning T do not love thee, Dr. Fell', &c. 

3 Or Mills, as the Oxford Dictionary of the Christian Church prefers, s.v. 

108 The Pre-critical Period 

Greek text, published exactly two weeks before his death at 
the age of sixty-two (23 June 1707). ■ Besides collecting all the 
evidence from Greek manuscripts, early versions, and Fathers 
that lay within his power to procure, Mill prefixed to his edition 
valuable prolegomena in which he dealt with the canon of the 
New Testament and the transmission of the New Testament 
text, described thirty-two printed editions of the Greek Testa- 
ment and nearly 100 manuscripts, and discussed patristic cita- 
tions from all the Fathers of any importance. Some idea of 
the extent and detail of the prolegomena may be had from the 
size of the index to the verses to which Mill makes reference in 
his discussion; these number 3,041 out of almost 8,000 verses 
in the whole New Testament. Despite the vast amount of solid 
learning embodied in his edition, however, Mill did not venture 
to form a text of his own, but reprinted Stephanus' text of 1 550 
without intentional variation. 

In 1 710 a reprint of Mill, with the prolegomena somewhat 
rearranged and with collations of twelve more manuscripts, was 
published at Amsterdam and Rotterdam by a Westphalian, 
Ludolf Kiister. Kiister's reprint also appeared, with a new title- 
page, at Leipzig in 1723 and again at Amsterdam in 1746. 

As Walton's critical efforts had been attacked by Owen, so 
also Mill's monumental work came under fire from the con- 
troversial writer, Dr. Daniel Whitby, Rector of St. Edmund's, 
Salisbury. Alarmed by the great number of variant readings 
which Mill had collected — some 30,000 in all — Whitby argued 
that the authority of the holy Scriptures was in peril, and that 
the assembling of critical evidence was tantamount to tamper- 
ing with the text. 2 

There were others, however, who, appreciating the textual 
evidence collected by Mill, attempted to embody in practical 

1 See Adam Fox, John Mill and Richard Bentley, a Study of the Textual Criticism of the 
New Testament, 16J5-1J29 (Oxford, 1954). 

2 Daniel Whitby, Examen variantium lectionum J. Millii (London, 1709). The 
English Deist, Anthony Collins (1 676-1 729), did, in fact, appeal to the existence of 
so many variant readings as an argument against the authority of the Scriptures {A 
Discourse of Freethinking [London, 1713]). The extent to which such considerations 
might be pushed is disclosed in Dean Swift's satirical essay, An Argument against the 
Abolition of Christianity, in which he refers to a roue 'who had heard of a text brought 
for proof of the Trinity, which in an ancient manuscript was differently read; he 
thereupon immediately took the hint, and by a sudden deduction of a long sorites, 
most logically concluded: "Why, if it is as you say, I may safely whore and drink 
on, and defy the parson" ' (Jonathan Swift, Works, iii [Edinburgh, 1814], p. 199). 

The Origin and Dominance of the Textus Receptus 109 

form the results of his honest critical endeavours. Between 1 709 
and 1719 Dr. Edward Wells (1667-1727), a mathematician and 
theological writer, published at Oxford, in ten parts, a Greek 
Testament with a variety of helps for the reader. 1 Wells deserted 
the Elzevir text 210 times, almost always agreeing with the 
judgement of nineteenth-century critical editors. Though Wells's 
edition was largely ignored by his contemporaries, history ac- 
cords him the honour of being the first to edit a complete New 
Testament which abandoned the Textus Receptus in favour of 
readings from the more ancient manuscripts. 

The name of Richard Bentley (1662- 1742), Master of Trinity 
College, Cambridge, is famous in the annals of classical scholar- 
ship for his exposure of the spurious Epistles of Phalaris, for his 
critical editions of Horace and Terence, for his discovery of the 
use of the digamma in the Homeric poems, and generally for his 
skill in textual emendation. At an early age Bentley began to 
correspond with various scholars on the subject of a critical 
edition of the Greek and Latin New Testament. In 1720 he 
issued a six-page prospectus of Proposals for Printing such an 
edition, giving as a specimen of his proposed text the last chapter 
of Revelation in Greek and Latin. 2 Here Bentley abandoned the 
Textus Receptus in more than forty places. 3 

By following the oldest manuscripts of the Greek original 
and of Jerome's Vulgate, Bentley was confident that he could 

1 The content of Wells's edition may be seen from, the lengthy descriptions on the 
title-pages, of which the following is a sample : 

An Help for the more Easy and Clear Understanding of the Holy Scriptures: 
being the Two Sacred Treatises of St Luke, Viz. his Gospel and the Acts of the 
Apostles, Explained after the following Method, viz. 

I. The Original or Greek Text amended, according to the Best and most Ancient 

II. The Common English Translation rendered more Agreeable to the Original. 

III. A Paraphrase, wherein not only the Difficult Expressions and Passages are 
explain'd, but also Each Treatise is divided into Proper Sections and Paragraphs: 
and withall it is observ'd, What Supplements to the Two Gospels of St Matthew 
and Mark are given us by St Luke in his Gospel. To the End of each Treatise is 
subjoin'd a Synopsis of the Contents therof. 

IV. Annotations relating (as Occasion requires) to the Several Particulars. 
Oxford, 1 719. 

2 For a reproduction of the pamphlet, see Caspar Rene' Gregory, Prolegomena 
(being vol. iii of Tischendorf 's Novum Testamentum Graece, ed. critica octava maior; 
Leipzig, 1884-94), PP- 231-40. 

3 Bentley's proposals were the occasion of an acrimonious controversy between 
Dr. Conyers Middleton and himself; for a summary of the points at issue, see James 
H. Monk, The Life of Richard Bentley, D.D., 2nd ed., ii (London, 1833), PP- : 3° ^ 

no The Pre-critical Period 

restore the text of the New Testament as it stood in the fourth 
century. 'By taking two thousand errors out of the Pope's 
Vulgate [Bentley refers to Pope Clement's edition of 1592], and 
as many out of the Protestant Pope Stephen's [referring to 
Stephanus' Greek text of 1550], I can set out an edition of each 
in columns, without using any book under nine hundred years 
old, that shall so exactly agree, word for word, and order for 
order, that no two tallies, nor two indentures, can agree better.' 1 
It is obvious that the Master of Trinity College was not inclined 
to underestimate his own abilities. In his Proposals he refers to the 
forthcoming edition as 'a Krij^a eaael, a charter, a Magna Ckarta, 
to the whole Christian church; to last when all the ancient 
MSS. here quoted may be lost and extinguished'. 

In order to finance the publication, subscriptions were soli- 
cited and about £2,000 collected from considerably more than 
1,000 prospective purchasers of the edition. Despite the elabo- 
rate plans, however, and the amassing of new evidence from 
manuscripts and Fathers, the scheme came to naught, and after 
Bentley's death his literary executor returned the money to the 
subscribers. 2 

While Bentley was gathering materials for a definitive edition 
which would supplant the Textus Receptus, a Greek and English 
diglot in two volumes was published anonymously at London 
in 1729 with the title, The New Testament in Greek and English. 
Containing the Original Text Corrected from the Authority of the Most 
Authentic Manuscripts: and a New Version Form'd agreeably to the 
Illustrations of the most Learned Commentators and Critics: with Notes 
and Various Readings, and a Copious Alphabetical Index. The edition 
has several typographical peculiarities. In the Greek text the 
editor discards smooth breathing marks and accents, and at the 
close of questions he uses '?' instead of the Greek mark of in- 
terrogation (the semicolon). In the English translation, as well 
as in the explanatory notes, he begins sentences with a capital 
letter only at the beginning of a new paragraph. 

The editor of this diglot edition was Daniel Mace, a Presby- 
terian minister at Newbury, who chose from Mill's apparatus 
those variant readings which seemed to him to be superior to the 
Textus Receptus. In a high proportion of these alterations Mace 

1 Bentleii Critica Sacra, ed. by A. A. Ellis (Cambridge, 1862), p. xv. 

2 See Fox, op. cit., pp. 105-26. 

The Origin and Dominance of the Textus Receptus 


anticipated the opinions of much later scholars. 1 Likewise his 
English translation reveals a certain independent vigour, for 
Mace adopted many racy and colloquial expressions; for 
example, 'don't', 'can't', 'what's', and (words of Simon the 
Pharisee to Jesus, Luke vii. 40) 'master, said he, lets hear it.' 
Here and there he anticipated modern versions ; for example, in 
Matt. vi. 27, instead of the King James reference to adding one 
cubit to one's stature, Mace renders, 'who by all his sollicitude 
can add one moment to his age?' (In Luke xii. 25 he translates, 
'but which of you, with all his disquietude, can add one moment 
to the period of his life?') In footnotes and appended notes Mace 
gives reasons for his departure from the earlier traditional text and 
translation. Several of these notes indicate Mace's free and inde- 
pendent spirit. Thus, in his note onZiva opos eaTivevrrjApa^ia (Gal. 
iv. 25), he declares: 'This has all themarks of an interpolation: it 
is quite foreign to the argument, and serves only to perplex the 
apostle's reasoning, which without it appears very clear and 
coherent.' He dismisses Mill's argument (Proleg. 1306), based on 
the unanimity of manuscript evidence in its favour, with the 
contemptuous remark, 'as if there was any manuscript so old as 
Common Sense' (p. 689), and prints the conjecture, to yap Ayap 
ovoToixei Tjj vuv IepovaaXrj[jL. ... In his extended discussion of the 
title of the Epistle to the Hebrews and the authorship of the 
Epistle, he writes: 'A very learned writer of our own thinks 
xiii. 23 a sufficient proof that Paul was the original author, as if 
no body could be acquainted with Timothy but Paul, which 
shows, that in order to understand the doctrine of Moral 
Evidence, that is, the doctrine of Chances, some other dis- 
cipline and diet is necessary besides that of bearly chewing a 
few Hebrew roots' (p. 840). 

Like the work of many other innovators, Mace's edition was 
either vehemently attacked or quietly ignored. In England 
Dr. Leonard Twells, Vicar of St. Mary's in Marlborough, 
issued in three parts A Critical Examination of the late New Testa- 
ment and Version of the New Testament: wherein the Editor's Corrupt 
Text, False Version, and fallacious Notes are Detected and Censured 
(London, 1731-2), 2 and on the Continent scholars like Pritius, 

1 For a list of some of these, see Eduard Reuss, Bibliotheca Novi Testamenti graeci . . . 
(Brunsvigae, 1872), pp. 175 f. 
1 In Part I under the heading, 'False Renderings, and other foul Management 

ii2 The Pre-critical Period 

Baumgarten, and Masch rivalled Twells in their invective and 
abuse of Mace. But most theologians assumed an ostrich-like 
pose, and Mace's work was soon all but forgotten. 1 

With Johann Albrecht Bengel (1687-1752) we reach a new 
stage in the history of the textual criticism of the New Testa- 
ment. While a student in theology at Tubingen his pietistic 
faith in the plenary inspiration of the Bible was disturbed by the 
30,000 variants which had recently been published in Mill's 
edition of the Greek Testament, and he resolved to devote him- 
self to the study of the transmission of the text. With charac- 
teristic energy and perseverance he procured all the editions, 
manuscripts, and early translations available to him. After ex- 
tended study he came to the conclusion that the variant readings 
were fewer in number than might have been expected, and that 
they did not shake any article of evangelic doctrine. 

In 1725, while teaching at the Lutheran preparatory school 
for ministerial candidates at Denkendorf, Bengel published an 
elaborate essay as a 'Forerunner' to his projected edition of the 
New Testament. 2 Here he laid down sound critical principles. 
He recognized that the witnesses to the text must not be counted 
but weighed, that is, classified in 'companies, families, tribes, 
nations'. He was accordingly the first to distinguish two great 
groups or 'nations' of manuscripts: the Asiatic, which originated 
from Constantinople and its environs and which included the 
manuscripts of more recent date ; and the African, which he 
subdivided into two tribes, represented respectively by codex 
Alexandrinus and the Old Latin. For the weighing of variant 
readings Bengel formulated a canon of criticism that, in one 
form or other, has been approved by all textual critics since. It is 
based on the recognition that a scribe is more likely to make a 
difficult construction easier, than make more difficult what was 
already easy. Formulated in Bengel's pithy Latin it is, proclivi 
scriptioni praestat ardua ('the difficult is to be preferred to the easy 

favouring Arianism' (pp. 1 34-44) , Twells lists fifteen examples of what he, in some 
cases with justice, regarded as biased translation. 

1 See now H. McLachlan's article, 'An Almost Forgotten Pioneer in New Testa- 
ment Criticism', Hibbert Journal, xxxvii (1938-9), pp. 617-25. 

2 This essay is entitled 'Prodromus Novi Testamenti recte cauteque ordinandi' 
and was included as an appendix to his edition of Chrysostomi libri VI de sacerdotio 
(Denkendorf, 1725). 

ii2 The Pre-critical Period 

Baumgarten, and Masch rivalled Twells in their invective and 
abuse of Mace. But most theologians assumed an ostrich-like 
pose, and Mace's work was soon all but forgotten. 1 

With Johann Albrecht Bengel (1687-1752) we reach a new 
stage in the history of the textual criticism of the New Testa- 
ment. While a student in theology at Tubingen his pietistic 
faith in the plenary inspiration of the Bible was disturbed by the 
30,000 variants which had recently been published in Mill's 
edition of the Greek Testament, and he resolved to devote him- 
self to the study of the transmission of the text. With charac- 
teristic energy and perseverance he procured all the editions, 
manuscripts, and early translations available to him. After ex- 
tended study he came to the conclusion that the variant readings 
were fewer in number than might have been expected, and that 
they did not shake any article of evangelic doctrine. 

In 1725, while teaching at the Lutheran preparatory school 
for ministerial candidates at Denkendorf, Bengel published an 
elaborate essay as a 'Forerunner' to his projected edition of the 
New Testament. 2 Here he laid down sound critical principles. 
He recognized that the witnesses to the text must not be counted 
but weighed, that is, classified in 'companies, families, tribes, 
nations'. He was accordingly the first to distinguish two great 
groups or 'nations' of manuscripts: the Asiatic, which originated 
from Constantinople and its environs and which included the 
manuscripts of more recent date; and the African, which he 
subdivided into two tribes, represented respectively by codex 
Alexandrinus and the Old Latin. For the weighing of variant 
readings Bengel formulated a canon of criticism that, in one 
form or other, has been approved by all textual critics since. It is 
based on the recognition that a scribe is more likely to make a 
difficult construction easier, than make more difficult what was 
already easy. Formulated in Bengel's pithy Latin it is, proclivi 
scriptioni praestat ardua ('the difficult is to be preferred to the easy 

favouring Arianism' (pp. 134-44), Twells lists fifteen examples of what he, in some 
cases with justice, regarded as biased translation. 

1 See now H. McLachlan's article, 'An Almost Forgotten Pioneer in New Testa- 
ment Criticism', Hibbert Journal, xxxvii (1938-9), pp. 617-25. 

2 This essay is entitled 'Prodromus Novi Testamenti recte cauteque ordinandi' 
and was included as an appendix to his edition of Chrysostomi libri VI de sacerdotio 
(Denkendorf, 1725). 

The Origin and Dominance of the Textus Receptus 113 

In 1734 Bengcl published at Tubingen an edition of the 
Greek New Testament in a handsome quarto volume. He did 
not venture to correct the traditional Textus Receptus in accord 
with personal judgement, but followed (except in nineteen pas- 
sages in the Book of Revelation) the self-imposed rule of not 
printing any reading which had not been previously published 
in an earlier printed edition. He indicated in the margin, how- 
ever, his views of the relative value of the variant readings 
according to the following categories : a designates the original 
reading ; /3 a reading which is better than that which is printed 
in the text; y a reading just as good as that in the text; 8 less good 
than the text; e a very inferior reading to be rejected. Bengel 
also took great pains to standardize the punctuation of the New 
Testament and to divide it into paragraphs — features which 
later editors borrowed from his edition. More than half of the 
volume is devoted to three excursuses in which Bengel supplied 
a reasoned account of his principles of textual criticism, and an 
apparatus drawn from Mill's collations plus his own collations 
of twelve additional manuscripts. 1 

Though Bengel was a man whose personal piety and life of 
good works were known to all (he had been in charge of an 
orphan home at Halle) and whose orthodoxy of belief was ac- 
knowledged (he was Superintendent of the Evangelical church 
of Wurttemberg) , he was treated as though he were an enemy of 
the holy Scriptures. So many persons impugned his motives and 
condemned his edition that he published in German, and then 
in Latin, a Defence of the New Testament. After BengePs death his 
son : in-law, Philip David Burk, published in 1 763 an enlarged 
edition of the apparatus criticus, along with several short pamph- 
lets which Bengel had written to explain and defend his mature 
views on the correct methods of recovering the earliest form of 
the text of the New Testament. 

Among those who collated manuscripts for Bentley was Jo- 
hann Jakob Wettstein (1 693-1 754), a native of Basle. His taste 
for textual criticism showed itself early; when ordained to the 
ministry at the age of twenty he delivered an address on variant 
readings in the New Testament. His textual studies, however, 

1 A summary of his text-critical principles is available also in the preface to his 
celebrated commentary on the New Testament entitled Gnomon Movi Testamenti 
(sect. viii). 

ii4 The Pre-critical Period 

were interpreted by some as preparations for denying the doc- 
trine of the divinity of Christ, and in 1 730 he was deposed from 
his pastorate and driven into exile. He secured a position as 
professor of philosophy and Hebrew in the Arminian college at 
Amsterdam (in succession to the celebrated Jean Leclerc) , and 
resumed his textual studies. 1 In 175 1-2 the fruits of forty years of 
research appeared in his publication at Amsterdam of a magni- 
ficent edition of the Greek New Testament in two folio volumes. 
Though he printed the Elzevir text, he indicated in the margin 
those readings which he himself held to be correct. In an appen- 
dix entitled 'Animadversiones et cautiones ad examen Varia- 
rum Lectionum N. T. necessariae', 2 Wettstein sets forth a good 
deal of sound advice; for example, he states that codices autempon- 
dere, non numero estimandi sunt (§ xviii fin., 'manuscripts must be 
evaluated by their weight, not by their number'). Despite his 
generally excellent theoretical views, Wettstein was somewhat 
haphazard in applying his rules. Furthermore, he came to 
advocate (largely, it seems, in opposition to Bengel) the quite 
untenable theory that all of the early Greek manuscripts have 
been contaminated by the Latin versions, and that consequently 
the later Greek manuscripts should be relied upon as preserving 
a more authentic text. 

Wettstein' s apparatus is the first in which the uncial manu- 
scripts are regularly denoted by capital Roman letters, and 
the minuscule manuscripts (including lectionaries) by Arabic 
numerals — a system which has continued to be used to the pre- 
sent time. In addition to the textual material, Wettstein's edition 
provides a thesaurus of quotations from Greek, Latin, and rab- 
binical authors, illustrating the usage of words and phrases of 
the New Testament. Though his critical judgement was not as 
sound as Bengel's, his passion for the study of manuscripts, which 
took him on extensive journeys, resulted in the collation or re- 
collation of about 100 manuscripts, and his commentary is still 
a valuable storehouse of classical, patristic, and rabbinical lore. 3 

1 On Wettstein's wide-ranging lectures under the general rubric of 'philosophy', 
see C. L. Hulbert-PowelPs biography, John James Wettstein, i6g^-iys4 (London, 
!938), pp- 196 f. 

2 Vol. ii, pp. 851-74, reproduced, with some condensation, from his anony- 
mously published treatise Prolegomena ad Novi Testamenti graeci editionem accuratissi- 
mam . . . (Amsterdam, 1730), pp. 165-201. 

3 In an appendix to his edition of the Greek Testament Wettstein published the 

The Origin and Dominance of the Textus Receptus 115 

Though he published no edition of the Greek Testament, 
Johann Salomo Semler (1725-91), often regarded as the father 
of German rationalism, made noteworthy contributions to the 
science of textual criticism by his reprint of Wettstein's Pro- 
legomena, with discerning comments of his own. 1 Adopting 
Bengel's system of classifying manuscripts by groups, Semler 
carried the process still further by assigning the origin of 
Bengel's Asiatic group (which he renamed Eastern) to the 
recension 2 prepared in the early fourth century by Lucian of 
Antioch, and the origin of Bengel's African group (Semler's 
Western or Egypto-Palestinian group) to Origen. Subsequently 
Semler expanded his textual researches and classified New Tes- 
tament manuscripts in three recensions: (1) Alexandrian, de- 
rived from Origen and preserved in the Syriac, Bohairic, and 
Ethiopic versions; (2) Eastern, current in the Antiochian and 
Constantinopolitan Churches ; and (3) Western, embodied in the 
Latin versions and Fathers. Later witnesses, he thought, were 
characterized by a mixture of all recensions. 3 

For the next important edition of the Greek Testament we 
must return to England, where William Bowyer, Jr. (1699- 
1777), was the third generation in a line of famous printers in 
London. Often regarded as 'the most learned English printer of 
whom we have any account', 4 Bowyer not only exercised a 
scholarly vigilance in printing a wide variety of volumes, but 
frequently contributed learned prefaces, annotations, and cor- 
rigenda to the works which passed through his publishing house. 
After his father and he had issued several editions of the Textus 
Receptus of the Greek Testament (in 1715, 1728, 1743, and 
1760), Bowyer decided to produce a critical edition worthy of 
the reputation of his firm. In 1763 such an edition was issued in 
two volumes of duodecimo size. Bowyer constructed his text 

editio princeps of the Syriac text of the two pseudo-Clementine epistles, De virginitate, 
with a Latin translation in parallel columns. 

1 J. S. Semler, Witstenii libelli adcrisin atque interpretationem Novi Testamenti (Halle, 

2 Semler was the first to apply the term 'recension' to groups of New Testament 
witnesses (Hermeneutische Vorbereitung, iii (i) [Halle, 1765]). Properly a recension is 
the result of deliberate critical work by an editor ; it is, however, often used in a 
loose sense as synonymous with 'family'. 

3 Semler, Apparatus ad liberalem Novi Testamenti interpretationem (Halle, 1767). 

4 S. Austin Allibone, A Critical Dictionary of English Literature, i (Philadelphia, 
1871), p. 229. 

n6 The Pre-critical Period 

largely by following the critical judgements which Wettstein 
had expressed in his marginal notes regarding the earliest form 
of text. By using square brackets Bowyer marked in his text not 
a few familiar passages which lack the support of good manu- 
scripts; for example, he bracketed the doxology of the Lord's 
Prayer (Matt. vi. i •$), the pericope de adultera( John vii. 53-viii. 11), 
the comma Johanneum ( 1 John v. 7-8) , and single verses (such as 
Acts viii. 37 and xv. 34) and words throughout the New Testa- 
ment. In other passages Bowyer departed from the Textus 
Receptus by introducing into his edition readings which the 
better manuscripts support. 1 In an appendix to the second 
volume Bowyer included nearly 200 pages of conjectural 
emendations on the New Testament, bearing on the text and 
punctuation of the New Testament. 2 

Another Englishman, Edward Harwood (1729-94), a Non- 
conformist minister, published at London in 1776 a two-volume 
edition of the New Testament which was, according to the 
statement on the title-page, 'Collated with the most approved 
Manuscripts; with Select Notes in English, critical and explana- 
tory ; and References to those Authors who have best illustrated 
the Sacred Writings. To which are added, a Catalogue of the 
principal Editions of the Greek Testament ; and a List of the 
most esteemed Commentators and critics.' For the Gospels and 
Acts Harwood followed in the main the text of codex Bezae, and 
for the Pauline letters codex Claromontanus. Where these were 
not available he utilized other manuscripts, chiefly codex 
Alexandrinus. 3 In an analysis of 1,000 passages in the New 

1 For a selected list of these, along with other details of Bowyer's edition, refer- 
ence may be made to the present writer's book, Chapters in the History of New Testa- 
ment Textual Criticism (Leiden and Grand Rapids, 1963), pp. 155-60. 

2 A second edition of the Conjectures, with extensive additions, was published 
separately at London in 1772 with the following title, Critical Conjectures and Observa- 
tions on the New Testament, Collected from Various Authors, as Well in Regard to Words as 
Pointing: With the Reasons on Which Both are Founded. This volume was translated into 
German by Joh. Chr. F. Schulz (2 vols., Leipzig, 1774-5). Two other editions, 
enlarged still further, were published posthumously at London, one in 1 782 and the 
other in 1812, the latter from the annotated interleaved copy of Dr. Henry Owen. 

3 In his preface Harwood declares, 'Excepting typographical errors, which a 
moderate acquaintance with the language will easily enable the reader to correct, I 
persuade myself, that the Text of the inspired writers here exhibited will approve 
itself to every Scholar who is a judge of sacred criticism, to be as near to the original 
autograph of the Evangelists and Apostles as any hitherto published in the world. 
To accomplish this arduous design, I carefully read through the late Professor 
Wetstein's Greek Testament, published at Amsterdam in two Volumes in folio, 

The Origin and Dominance of the Textus Receptus 117 

Testament, Reuss found that Harwood deserted the Textus 
Receptus more than 70 per cent, of the time, and in 643 passages 
agrees with the epoch-making critical edition of Lachmann, 
published in the nineteenth century (see pp. 124-6 below). 
It is not surprising that, in a period when the Textus Receptus 
held sway and when only occasionally an independent spirit 
ventured to question its authority, the first Greek Testament to 
be published in America was the time-honoured Textus Re- 
ceptus. 1 The printer of this edition, Isaiah Thomas, Jr. (1749- 
1831), was a typically enterprising and hard-working Yankee. 
Apprenticed to a printer at the age of six, after only six weeks 
of indifferent schooling, Thomas began his upward climb that 
resulted in his becoming a member of nearly every learned 
society in the country and the recipient of honorary degrees 
from Dartmouth and Allegheny colleges. His printing establish- 
ment issued over 900 different books, more than those of 
Benjamin Franklin, Hugh Gaine, and Matthew Carey, his 
nearest rivals. Indeed, owing to the excellence of his typo- 
graphical work and the range and number of his imprints, 
Franklin called him 'the Baskerville of America'. Sensing that 
the market called for an edition of the Greek New Testament, 
Thomas secured the assistance of a scholarly minister, the Rev. 
Caleb Alexander, 2 and issued the editio prima Americana of the 
Greek New Testament. The volume, a small-sized duodecimo 
of 478 pages, was published at Worcester, Massachusetts, in 
April 1800. 

scrupulously weighed the merit or demerit of the various lections there exhibited 
from a great multitude of Manuscripts of different value, and adopted only those 
which to my judgment appeared to be best authenticated: my meaning is, that I 
espoused only those which I verily believed to be the very words which the inspired 
authors originally wrote' (pp. viii-ix). 

1 The first Bible to be printed in America was the translation made by John 
Eliot into the Algonquin Indian language, published at Cambridge, Massachusetts, 
in 1661^3. The first Bible in a European language to be printed in America was 
Luther's German translation published in 1743 by Christoph German- 
town, Pennsylvania. The first Bible in English to be published in America came 
from the press of Robert Aitken at Philadelphia in 1782. 

2 Caleb Alexander (1755-1828), a native of Northfield, Massachusetts, was 
graduated from Yale College in 1777. During his pastorate at Mendon, a village 
not far from Worcester, Alexander found time to write two Latin grammars and a 
Greek grammar, the latter being published by Thomas with the title Grammatical 
System of the Grecian Language (Worcester, 1 796) . It was doubtless at this period that 
Thomas secured Alexander's services in supervising the preparation of a Greek New 

1 1 8 Pre-critical Period — Textus Receptus 

The title-page states that the edition reproduces accurately 
the text of John Mill's edition (juxta exemplar Joannis Millii 
accuratissime impressum). This is, however, not entirely true to 
fact, for in more than a score of passages the editorial work of 
Alexander is discernible. According to Isaac H. Hall, 1 a com- 
parison with editions issued by Beza and by the Elzevirs shows 
that Alexander, in eclectic fashion, occasionally chose a reading 
now from this edition and now from that. Externally the format 
of the volume bears many resemblances to Bowyer's editions of 
the Textus Receptus. In fact, the title-page of the Alexander- 
Thomas edition reproduces exactly, line for line, word for word, 
and style for style of type (except only as to date and names and 
place of publication), the title-page of Bowyer's edition of 1794. 2 

1 Isaac H. Hall, A Critical Bibliography of the Greek New Testament as Published in 
America (Philadelphia, 1883), P- "■ 

2 For further information regarding the publisher and the text of this first 
American edition of the Greek Testament, see B. M. Metzger, 'Three Learned 
Printers and their Unsung Contributions to Biblical Scholarship', Journal of Religion, 
xxxii (1952), pp. 254-62. 


The Modern Critical Period: 
From Griesbach to the Present 


During the latter part of the eighteenth century the Ger- 
man scholar Johann Jakob Griesbach (1745-18 12) laid 
foundations for all subsequent work on the Greek text 
of the New Testament. A pupil of Semler's at Halle, Griesbach 
was Professor of New Testament at the University of Jena from 
1775 un til his death. After travelling in England, Holland, and 
France in order to collate manuscripts, he devoted special at- 
tention to the New Testament quotations in the Greek Fathers 
and to several versions of the New Testament which previously 
had been little studied, such as the Gothic, the Armenian, and 
the Philoxenian Syriac. 

Griesbach also investigated the history of the transmission of 
the New Testament text in antiquity and further developed 
Bengel's and Semler's grouping of manuscripts in recensions. 
At first he was inclined to divide the extant materials into five 
or six different groups ; he afterwards limited them to three, the 
Alexandrian, Western, and Byzantine recensions. The standard 
of the Alexandrian text he believed to be Origen, who, though 
writing many of his works in Palestine, was assumed to have 
brought with him into exile copies of the Scriptures similar to 
those used in his native city. To this group Griesbach assigned 
the uncial manuscripts C, L, and K, the minuscules 1,13, 33, 69, 
106, and 118, the Bohairic, Armenian, Ethiopic, and Harclean 
Syriac, and in addition to quotations of Origen, those of Cle- 
ment of Alexandria, Eusebius, Cyril of Alexandria, and Isidore 
of Pelusium. To the Western group he assigned codex D, the 
Latin versions, and, in part, the Peshitta Syriac and the Arabic 
versions. The Constantinopolitan group, which he regarded as 
a later compilation from the other two, was represented by A 

120 The Modern Critical Period 

(in the Gospels), and by the great mass of later uncial and minus- 
cule manuscripts, as well as the larger proportion of patristic 

Among the fifteen canons of textual criticism which Griesbach 
elaborated, the following (his first canon) may be given as a 
specimen : 

The shorter reading (unless it lacks entirely the authority of the 
ancient and weighty witnesses) is to be preferred to the more verbose, 
for scribes were much more prone to add than to omit. They scarcely 
ever deliberately omitted anything, but they added many things; 
certainly they omitted some things by accident, but likewise not a 
few things have been added to the text by scribes through errors of 
the eye, ear, memory, imagination, and judgement. Particularly the 
shorter reading is to be preferred, even though according to the 
authority of the witnesses it may appear to be inferior to the other, — 

(a) if at the same time it is more difficult, more obscure, ambiguous, 
elliptical, hebraizing, or solecistic; 

(b) if the same thing is expressed with different phrases in various 

(c) if the order of words varies ; 

(d) at the beginning of pericopes; 

(e) if the longer reading savours of a gloss or interpretation, or 
agrees with the wording of parallel passages, or seems to have come 
from lectionaries. 

But on the other hand the longer reading is to be preferred to the 
shorter (unless the latter appears in many good witnesses), — 

(a) if the occasion of the omission can be attributed to homoeo- 
teleuton ; 

(b) if that which was omitted could have seemed to the scribe to 
be obscure, harsh, superfluous, unusual, paradoxical, offensive to 
pious ears, erroneous, or in opposition to parallel passages; 

(c) if that which is lacking could be lacking without harming the 
sense or the structure of the sentence, as for example incidental, brief 
propositions, and other matter the absence of which would be 
scarcely noticed by the scribe when re-reading what he had written ; 

(d) if the shorter reading is less in accord with the character, style, 
or scope of the author; 

(e) if the shorter reading utterly lacks sense; 

(/) if it is probable that the shorter reading has crept in from 
parallel passages or from lectionaries. 

From Griesbach to the Present 121 

Griesbach showed great skill and tact in evaluating the evi- 
dence of variant readings. For example, his judgement, based 
on patristic and versional evidence, that the shorter form of the 
Lord's Prayer in Luke xi. 3-4 is to be preferred was remarkably 
confirmed a few years later when the readings of codex Vatica- 
nus were published, for it was found that all of the omissions are 
supported by that early manuscript. 

The importance of Griesbach for New Testament textual 
criticism can scarcely be overestimated. For the first time in 
Germany a scholar ventured to abandon the Textus Receptus 
at many places and to print the text of the New Testament in 
the form to which his investigations had brought him. Though 
at times Griesbach permitted himself to be led astray by a too 
mechanical adherence to his system of recensions, 1 his textual 
labours on the whole were characterized by caution and can- 
dour. His principal editions were published at Halle in 1775—7, 
at Halle and London in 1796-1806, and at Leipzig in 1803-7. 
Several editions of his text were also issued by enterprising 
printers in England, Scotland, and America. His influence was 
extended still further when his work was adopted as the basis 
of smaller manual editions issued on the Continent by Schott, 
Knapp, Tittmann, Hahn, and Theile. 

Soon after the publication of Griesbach's first edition, several 
other scholars published collations which greatly increased the 
availability of evidence for the text of the New Testament from 
Greek manuscripts, the early versions, and the Church Fathers. 
Christian Friedrich Matthaei (1744-18 11), professor first at 
Wittenberg and then at Moscow, where he taught classical 
literature, issued at Riga in twelve parts, between 1782 and 
1 788, an edition of the Greek text with the Latin Vulgate. His 
printed text is of little value because it is based on manuscripts 
of recent date, but his apparatus is valuable. Besides collat- 
ing manuscripts at Dresden, Leipzig, and Gottingen, Matthaei 
sought out Biblical and patristic manuscripts in Moscow, ori- 
ginally brought to Russia from Mount Athos. 2 He made, for 

1 He was vehemently criticized for this and other reasons by Archbishop R. 
Laurence in his Remarks on the Systematical Classification of Manuscripts adopted by 
Griesbach in his Edition of the New Testament (Oxford, 1814; reprinted in the Biblical 
Repertory, edited by Charles Hodge, ii [1826], pp. 33-95). See below, p. 270. 

2 While in Russia Matthaei managed to steal a good many manuscripts of both 
the classics and the Fathers. Some of these he kept in his own library, while others 

122 The Modern Critical Period 

example, collations of thirty-four manuscripts of the homilies 
of John Chrysostom on the Gospels and the Pauline Epistles. In 
the second edition of his New Testament, without the Latin Vul- 
gate (3 vols., 1803-7), Matthaei provides evidence from about 
thirty additional manuscripts. His edition is noteworthy as con- 
taining, apparently for the first time, evidence from the Slavic 
version of the New Testament. In an appendix to his edition of 
the Book of Revelation Matthaei lists ten Slavic manuscripts 
which he had examined; he contented himself, however, with 
collating the text of Revelation in the folio edition of the Slavic 
Bible published at Moscow in 1762. This evidence is given in 
Latin, the collation having been made against a manuscript of 
the Latin Vulgate at Moscow, codex Demidovianus. 

Franz Karl Alter (1749- 1804), a Jesuit of Silesia who be- 
came Professor of Greek at Vienna, published an edition of the 
Greek Testament in two volumes (Vienna, 1 786-7) , based on the 
text of a single manuscript in the Imperial Library at Vienna. 
In separate appendixes he cited evidence from twenty other 
Greek manuscripts, two Latin manuscripts, the Bohairic version 
(which David Wilkins had edited at Oxford in 17 16), and four 
Slavic manuscripts. This is the first edition of the Greek New 
Testament which contains evidence from Slavic manuscripts 

A still larger addition to the quantity of materials available to 
the textual critic was made through the efforts of four Danish 
scholars, Andreas Birch, Jacob G. C. Adler, D. G. Molden- 
hauer, and O. G. Tychsen, who were sent by the King of 
Denmark, Christian VII, to examine manuscripts in Italy, 
Spain, Germany, and other European countries. The results of 
their labours were published by Birch in a volume describing the 
then known Greek manuscripts of the New Testament, 1 and in 
four volumes of collations (Copenhagen, 1 788-1801). The latter 
contain variant readings from 172 Greek manuscripts, and also 
evidence from two Syriac versions (the Philoxenian and Pales- 
tinian) . Many of the manuscripts, however, were only partially 

he sold or gave to various libraries and friends in Germany and Holland. For an 
account of his life with incriminating evidence of his brazen thievery, see Oscar 
von Gebhardt in Centralblatt fur Bibliothekswesen, xv (1898), pp. 345-57, 393 _ 420, 
441-82, and 537-66. 

1 Andreas Birch, Krilisk Beskrivelse over graske Haandskrifter af det Nye Testamente 
(Copenhagen, 1785). 

From Griesback to the Present 123 

examined by Birch and his colleagues, including codex Vati- 
canus (B), readings of which now for the first time appeared in 

About this time two Roman Catholic scholars gave, in differ- 
ent ways, an impetus to the textual criticism of the New Testa- 
ment. Johann Leonhard Hug (1 765-1 846), professor at the 
University of Freiburg in Breisgau, developed the theory 1 that 
at about the beginning of the third century the several types of 
New Testament text degenerated rapidly and produced what is 
commonly called the Western text, which Hug called the Koivfj 
eicSoais (the common edition). Toward the middle of the third 
century, according to Hug, this edition was revised in Palestine 
by Origen, a revision adopted later by Jerome ; in Egypt it was 
revised by Hesychius, and in Syria by Lucian, a presbyter of 
Antioch, both of which revisions Jerome condemned. Although 
Hug started from what was on the whole a true conception of 
the Western text and its manifold variations, his ingenious 
attempt to connect three recensions of the Septuagint (whose 
places of origin he believed assured) with three types of New 
Testament text failed. 

Johannes Martin Augustinus Scholz (1794-1852), a pupil of 
Hug's and professor at the University of Bonn, travelled exten- 
sively throughout Europe and the Near East in order to draw up 
what was the first comprehensive listing of Greek manuscripts 
of the New Testament, adding 616 new manuscripts to those 
previously known. He was the first to emphasize the importance 
of ascertaining the geographical provenance represented by the 
several manuscripts, a point which B. H. Streeter was to elabor- 
ate in 1924 by his theory of 'local texts'. Unlike Streeter, who 
relied on the congruence of manuscript readings with patristic 
citations, Scholz was guided chiefly by certain external signs 
of provenance, such as details of palaeography, iconography, 
marginal notes, colophons, and evidence regarding local saints 
who were honoured in lectionaries. 

After some tentative attempts at classifying manuscripts, 
Scholz came to adopt essentially Bengel's division into two 
families, which he called the Alexandrian and the Constan- 
tinopolitan. During his extensive examinations of minuscule 

1 J. L. Hug, Einleitung in die Schriften des Neuen Testaments (Stuttgart, 1808), 

i24 The Modern Critical Period 

manuscripts he was impressed by their general uniformity of 
text-type, a feature which he regarded as evidence of their 
superiority to the earlier Alexandrian type. Thus Scholz's 
two- volume edition of the Greek Testament (Leipzig, 1830-6) 
marked a retrogression in textual criticism toward the Textus 
Receptus ; only here and there does it happen to contain read- 
ings supported by the earlier manuscripts, because the editor 
was inconsistent in the application of his critical theories. It is 
symptomatic of the low ebb to which appreciation of textual 
scholarship had sunk in England at this time that Scholz's edi- 
tion was welcomed and praised by many British scholars, and 
its text was reprinted by Bagster in London in several editions. 
At a later date (1845) Scholz retracted his preference for the 
Byzantine text, and declared that, if a new edition of his Greek 
Testament were called for, he would receive into the text most of 
the Alexandrian readings which he had formerly placed in the 


The first recognized scholar to break totally with the Tex- 
tus Receptus was the celebrated classical and Germanic philo- 
logist of Berlin, Karl Lachmann (1793-1851), who published 
an edition of the Greek Testament which rests wholly upon the 
application of textual criticism in the evaluation of variant 
readings. Lachmann is famous for his editions of ancient classical 
authors, including Propertius, Catullus, Tibullus, Lucretius, as 
well as medieval epics and lyrics, such as the Nibelungenlied, 
Walther von der Vogelweide, and Wolfram von Eschenbach. 
He demonstrated how, by comparison of manuscripts, it is 
possible to draw inferences as to their lost ancestors or arche- 
types, their condition, and their pagination. In his most famous 
work, that on Lucretius, he showed that the peculiarities of the 
three chief manuscripts all derive from a single archetype, con- 
taining 302 pages of twenty-six lines to the page, and thus he 
was enabled to make various transpositions in the received text. 

In editing the New Testament Lachmann's aim was not to 
reproduce the original text, which he believed to be an impos- 
sible task, but to present on purely documentary evidence, apart 
from any previously printed editions, the text current in Eastern 

From Griesbach to the Present 125 

Christendom at the end of the fourth century (about A.D. 380). 
Using no minuscule manuscripts, he based his text on several of 
the earlier uncials, the Old Latin and Jerome's Vulgate, and 
the testimony of Irenaeus, Origen, Cyprian, Hilary, and Luci- 
fer. After five years of work, in 1831 he published at Berlin an 
edition of the Greek text, with a list of passages where it differs 
from the Textus Receptus. Brackets are used to indicate words 
of doubtful textual authority. Instead of including in the edition 
itself an account of his methodology and the reasons that led 
him to reject the Textus Receptus, Lachmann chose to refer 
the reader to an article which he had published the previous 
year in a German periodical. 1 It is not surprising that theo- 
logians, even liberal ones like de Wette, generally misunderstood 
Lachmann's intentions and attacked him with considerable 
vehemence, coining such names of reproach as 'Bentley's ape' 
(simia Bentleii). In the preface to his second edition (2 vols., 
Berlin, 1842-50), Lachmann replied in kind, arrogantly twit- 
ting his critics for their blind preference for the familiar but 
corrupt later text to the earlier, purer form. 

It was not always appreciated that Lachmann did not pretend 
to print the original text of the New Testament, but only a pro- 
visional one, namely that current in the fourth century, in- 
cluding even palpable scribal errors if sufficiently well attested. 
The weakness of the edition is the slender manuscript basis to 
which Lachmann restricted himself. According to Scrivener, 
'Lachmann's text seldom rests on more than four Greek codices, 
very often on three, not infrequently on two; in Matt. vi. 20- 
viii. 5, and in 165 out of 405 verses of the Apocalypse, on but 
one.' z Despite such limitations, however, the judgement of most 
later scholars has agreed with Hort's evaluation of Lachmann 
and his work: 

A new period began in 1831, when for the first time a text was 
constructed directly from the ancient documents without the inter- 
vention of any printed edition, and when the first systematic attempt 
was made to substitute scientific method for arbitrary choice in the 
discrimination of various readings. In both respects the editor, 

1 'Rechcnschaft iiber seine Ausgabe des Neuen Testaments', Theologische Studien 

und Kritiken, [iii] (1830), pp. 817-45. 

2 F. H. A. Scrivener, A Plain Introduction to the Criticism of the New Testament 
4th ed., ii (London, 1894), p. 233. 

126 The Modern Critical Period 

Lachmann, rejoiced to declare that he was carrying out the prin- 
ciples and unfulfilled intentions of Bentley, as set forth in 1 7 1 6 and 
1720. • 

The man to whom modern textual critics of the New Testa- 
ment owe most is without doubt Lobegott Friedrich Constantin 
von Tischendorf (181 5-74) , who sought out and published more 
manuscripts and produced more critical editions of the Greek 
Bible than any other single scholar. Between 1841 and 1872 he 
prepared eight editions of the Greek Testament, some of which 
were reissued alone or with German or Latin versions, as well as 
twenty-two volumes of texts of Biblical manuscripts. The total 
number of his books and articles, most of them relating to 
Biblical criticism, exceeds 150. 2 

While studying theology at Leipzig from 1834 to 1838 young 
Tischendorf came under the influence of Johann G. B. Winer, 
whose grammar of New Testament Greek (1822) went through 
many editions and remained the standard for several genera- 
tions. Winer infused in his student a passion to seek and to 
utilize the most ancient witnesses in reconstructing the purest 
form of the Greek Scriptures. To this task the young scholar 
dedicated himself; writing to his fiancee, he declared: 'I am 
confronted with a sacred task, the struggle to regain the original 
form of the New Testament.' At the age of twenty-five, sup- 
ported by a small stipend from the Government of Saxony, 
Tischendorf began the laborious work of deciphering the 
palimpsest codex Ephraemi and certain other manuscripts in 
the Bibliotheque Nationale at Paris. Subsequently he visited 
libraries throughout Europe and the Near East searching for and 
examining manuscripts new and old. (For the story of his dis- 
covery of codex Sinaiticus, see pp. 42-45 above.) 

Of Tischendorf 's several editions of the Greek Testament the 
most important is the eighth (editio octava critica maior), issued in 
eleven parts, beginning in 1864, and published in two volumes 
(Leipzig, 1869-72). This was accompanied by a rich apparatus 
criticus in which Tischendorf assembled all of the variant read- 

1 Westcott and Hort, The New Testament in the Original Greek, [ii,] Introduction [and] 
Appendix (Cambridge, 1881), p. 13. See below, p. 270. 

2 See the biographical article on Tischendorf written by Caspar Rene Gregory 
in Bibliotheca Sacra, xxxiii (1876), pp. 153-93, containing a list of Tischendorf 's 

From Griesbach to the Present 127 

ings which he or his predecessors had found in manuscripts, 
versions, and Fathers. Soon after the publication of the second 
volume a stroke of palsy prevented Tischendorf from continuing 
his labours. A third volume of valuable Prolegomena to the edition 
was prepared by Caspar Rene Gregory and issued in three parts 
(Leipzig, 1884, 1890, 1894). 1 

Tischendorf's claim to fame rests chiefly upon his inde- 
fatigable industry in assembling textual evidence ; his use of that 
evidence in constructing his editions, however, was marked by 
a somewhat wooden adherence to a number of critical canons, 
as well as a certain arbitrariness in dealing with problems not 
covered by the canons. The text of his eighth edition differs 
(according to Eberhard Nestle) from the seventh edition in 
3,572 places, and he has been accused of giving excessive weight 
to the evidence of codex Sinaiticus, which he had discovered 
between issuing the two editions. 

In England the scholar who, at the middle of the nineteenth 
century, was most successful in drawing British preference away 
from the Textus Receptus was Samuel Prideaux Tregelles 
(1813-75). As a boy he had shown exceptional talent and in- 
tellectual curiosity, and while earning his livelihood at an 
ironworks managed to devote his spare time to learning Greek, 
Aramaic, Hebrew, and Welsh. While still in his early twenties, 
Tregelles began to form plans for a new critical edition of the 
New Testament. Having observed how persistently Scholz 
rejected the evidence of the earliest manuscripts, and being 
dissatisfied with the somewhat hesitating way in which Gries- 
bach still clung to the Textus Receptus, he determined to em- 
ploy his leisure time in preparing an edition based only on 
the evidence of the earliest witnesses. Without his knowing it, 
Tregelles developed critical principles that paralleled to a re- 
markable degree those of Lachmann. Thereafter he was engaged 
in the collation of Greek manuscripts, making extensive travels 
throughout Europe for the purpose. His careful and systematic 
examination of practically all the then known uncials and 
several of the important minuscules resulted in the correction of 
many erroneous citations made by previous editors. He also 

1 Gregory's volume of Prolegomena, with additions and corrections, was later 
published in German in three parts, entitled Textkritik des Neuen Testamentes (Leipzig, 

128 - The Modern Critical Period 

examined afresh the New Testament quotations found in Greek 
Church Fathers down to Eusebius, as well as the ancient versions, 
and edited (1861) a palimpsest manuscript of the Gospel of 
Luke, codex Zacynthius (3), acquired in 1821 by the British 
and Foreign Bible Society. Before issuing any portion of his 
new text, however, Tregelles published a survey of earlier edi- 
tions in. which he set forth his own critical principles (An Account 
of the Printed Text of the Greek Mew Testament . . ., London, 1854) ; 
he also rewrote that portion of T. H. Home's encyclopedic Intro- 
duction to the Critical Study and Knowledge of the Holy Scriptures 
which relates to the textual criticism of the New Testament 
(vol. iv, 10th ed., London, 1856). 

Unlike Tischendorf, who hurried into print with another edi- 
tion as soon as he had discovered some new manuscript evidence, 
Tregelles preferred to fix his full energy upon the final goal 
of a definitive text representing his mature judgement, and is- 
sued but one edition. This was published at London in six parts 
between 1857 and 1872. Disabled by a stroke of paralysis in 
1870, he secured the assistance of B. W. Newton for the final 
fascicle. A volume of prolegomena compiled from Tregelles's 
other works, and containing many pages of addenda et corrigenda, 
was edited by F. J. A. Hort and A. W. Streane and published 
posthumously in 1879. m spite of poverty, opposition, and ill 
health, Tregelles overcame all difficulties and devoted a life- 
time of meticulous labours upon the text of the New Testament 
as an act of worship, undertaken, as he declares in the preface, 
'in the full belief that it would be for the service of God, by 
serving His Church'. 

Though remembered primarily for his widely used com- 
mentary on the New Testament, Henry Alford (1810-71), 
Dean of Canterbury and author of several well-known hymns 
(among them 'Come, ye thankful people, come' and 'Ten 
thousand times ten thousand'), deserves mention here as an 
ardent advocate of the critical principles formulated by those 
who, like Lachmann, had worked for the 'demolition of the 
unworthy and pedantic reverence for the received text, which 
stood in the way of all chance of discovering the genuine word 
of God'. 1 In the successive editions of his commentary, Alford 

1 Henry Alford, The Greek Testament, with a critically revised Text . . ., new ed., i, 
(New York, 1881), p. 76 of the prolegomena. 

From Gnesbach to the Present 129 

set forth more and more fully the evidence of variant readings, 
and boldly printed that form of Greek text which he believed 
was supported by the earliest and best witnesses. 

The year 1881 was marked by the publication of the most 
noteworthy critical edition of the Greek Testament ever pro- 
duced by British scholarship. After working about twenty-eight 
years on this edition (from about 1853 to 1881), Brooke Foss 
Westcott (1825-1901), Canon of Peterborough and Regius Pro- 
fessor of Divinity at Cambridge (he was consecrated Bishop of 
Durham in 1890), and Fenton John Anthony Hort (1828-92), 
Hulsean Professor of Divinity at Cambridge, issued two volumes 
entitled, The New Testament in the Original Greek. Volume i 
contains the Greek text; volume ii comprises a valuable Intro- 
duction and Appendix, in which the critical principles followed by 
the two editors are set forth in detail by Hort, with the con- 
currence of his colleague, and certain problem passages are 
discussed. 1 

Unlike earlier editors, neither Westcott nor Hort was con- 
cerned to collate manuscripts, nor did they provide a critical 
apparatus. Rather, utilizing previous collections of variant 
readings, they refined the critical methodology developed by 
Griesbach, Lachmann, and others, and applied it rigorously, 
but with discrimination, to the witnesses to the text of the New 
Testament. The principles and procedures of criticism which 
they elaborated may be summarized as follows. 

Hort begins the classic Introduction by discussing what he calls 
Internal Evidence of Readings. 

The most rudimentary form of criticism [he writes] consists 
in dealing with each variation independently, and adopting at 
once in each case out of two or more variants that which looks 
most probable. . . . Internal Evidence of Readings is of two 
kinds, which cannot be too sharply distinguished from each other; 
appealing respectively to Intrinsic Probability, having reference 
to the author, and what may be called Transcriptional Prob- 
ability, having reference to the copyists. In appealing to the first, 
we ask what an author is likely to have written : in appealing to the 

1 Occasionally, when the editors could not agree in certain details, the opinion 
of each is identified by his initials. The second edition of the second volume, 
published in 1896, contains some additional notes by F. C. Burkitt on the recently 
discovered Sinaitic Syriac manuscript. 

130 The Modern Critical Period 

second, we ask what copyists are likely to have made him seem to 
write. 1 

When, as sometimes happens, Intrinsic and Transcriptional 
Probabilities are in conflict, it is usually safer to make judge- 
ments on the basis of what Hort called the 'observed proclivities 
of average copyists' than on what one imagines the original 
author must have written. 

In order to transcend the limitations inherent in a procedure 
based solely on Internal Evidence of Readings, the textual 
critic must also utilize Internal Evidence of Documents. When 
weighing the evidence in individual cases, one gains assurance 
by considering whether a witness is normally credible and trust- 
worthy. Therefore, instead of being content with evaluating one 
reading after another, in isolation from each other, the critic 
should collect information regarding the character of individual 
manuscripts. If one finds that a given manuscript frequently 
supports certain readings which clearly commend themselves 
as original on the basis of probability, it is natural to prefer 
its readings in other instances when the Internal Evidence of 
Readings is not clear enough for a decision. Hort summarizes 
this point by enunciating the principle that 'knowledge of docu- 
ments should precede final judgement upon readings'. 2 

The next step involves the examination of the relationship 
of the several witnesses to one another. Manuscripts may be 
grouped and considered from the standpoint of their genealogy. 
If, for example, of ten manuscripts nine agree against one, but 
the nine have a common original, the numerical preponderance 
counts for nothing. The clearest evidence in tracing the genea- 
logy of witnesses is the presence of conflate readings, that is, 
readings which have arisen from the combination of elements 
which had existed previously in separate manuscripts. Here 
Hort enunciates another principle of criticism, that 'all trust- 
worthy restoration of corrupted texts is founded on the study of 
their history, that is, of the relations of descent or affinity which 
connect the several documents'. 3 

Finally, in his discussion of methodology, Hort considers the 
Internal Evidence of Groups, which is in some sense inter- 

1 Westcott and Hort, op. cit., pp. ig— 20. 
* Ibid., p. 31. 
3 Ibid., p. 40. 

From Griesbach to the Present 131 

mediate between Internal Evidence of Documents and Genea- 
logical Evidence. Just as it is useful to determine the general 
characteristics of a given manuscript by observing how often it 
supports or rejects readings which have been previously evalu- 
ated individually on the basis of Internal Probability, so the 
general characteristics of a given group of witnesses can be 
determined and evaluated in relation to other groups. 

The validity of inferences based on this procedure depends on 
the genealogical principle that 'community of reading implies 
community of origin'. 1 Such generalizations on the value of 
groups of witnesses, in turn, assist the critic in coming to de- 
cisions in instances where mixture in the ancestry of manuscripts 
makes it difficult to draw up a genealogical family tree. 

The paragraphs above contain a summary of the critical 
principles adopted and elaborated by Westcott and Hort. The 
results of their application of these principles to the then known 
New Testament manuscripts will now be briefly set forth. 

On the basis of investigations as to relationships among the 
witnesses to the text of the New Testament, Westcott and Hort 
distinguished four principal types of text, which they called the 
Syrian, the Western, the Alexandrian, and the Neutral. 

(1) The latest of these four forms of text is the Syrian, which 
is a mixed text resulting from a revision made by an editor or 
editors in the fourth century who wished to produce a smooth, 
easy, and complete text. This conflated text, the farthest re- 
moved from the originals, was taken to Constantinople, whence 
it was disseminated widely throughout the Byzantine Empire* 
It is best represented today by codex Alexandrinus (in the Gos- 
pels, not in Acts and the Epistles) , the later uncial manuscripts, 
and the great mass of minuscule manuscripts. The Textus Re- 
ceptus is the latest form of the Syrian text. 

Hort's classic description of the Syrian text is as follows : 

The qualities which the authors of the Syrian text seem to have 
most desired to impress on it are lucidity and completeness. They 
were evidently anxious to remove all stumbling-blocks out of the way 
of the ordinary reader, so far as this could be done without recourse 
to violent measures. They were apparently equally desirous that he 
should have the benefit of instructive matter contained in all the 
existing texts, provided it did not confuse the context or introduce 

1 Ibid., p. 60. 

132 The Modern Critical Period 

seeming contradictions. New omissions accordingly are rare, and 
where they occur are usually found to contribute to apparent sim- 
plicity. New interpolations on the other hand are abundant, most 
of them being due to harmonistic or other assimilation, fortunately 
capricious and incomplete. Both in matter and in diction the Syrian 
text is conspicuously a full text. It delights in pronouns, conjunc- 
tions, and expletives and supplied links of all kinds, as well as in more 
considerable additions. As distinguished from the bold vigour of the 
'Western' scribes, and the refined scholarship of the Alexandrians, 
the spirit of its own corrections is at once sensible and feeble. Entirely 
blameless on either literary or religious grounds as regards vulgarised 
or unworthy diction, yet shewing no marks of either critical or spi- 
ritual insight, it presents the New Testament in a form smooth and 
attractive, but appreciably impoverished in sense and force, more 
fitted for cursory perusal or recitation than for repeated and diligent 
study. 1 

(2) Of the remaining types of texts which Westcott and Hort 
isolated, the so-called Western type is both ancient and wide- 
spread. It is preserved in certain bilingual uncial manuscripts, 
notably codex Bezae of the Gospels and Acts (D) and codex 
Claromontanus of the Epistles (D p ), in the Old Latin version(s), 
and in the Curetonian Syriac. Its date of origin must have been 
extremely early, perhaps before the middle of the second cen- 
tury. Marcion, Tatian, Justin, Irenaeus, Hippolytus, Tertul- 
lian, and Cyprian all made use to a greater or less extent of a 
Western form of text. 

One of the marked characteristics of the Western text, accord- 
ing to Hort, is a love of paraphrase : 

Words, clauses, and even whole sentences were changed, omitted, 
and inserted with astonishing freedom, wherever it seemed that the 

meaning could be brought out with greater force and definiteness 

Another equally important characteristic is a disposition to enrich 
the text at the cost of its purity by alterations or additions taken from 
traditional and perhaps from apocryphal or other non-biblical 
sources. [The Western text is also characterized by] the multipli- 
cation of genitive pronouns, but occasionally their suppression 
where they appeared cumbrous; the insertion of objects, genitive, 
dative, or accusative, after verbs used absolutely, the insertion of 
conjunctions in sentences which had none, but occasionally their 
excision where their force was not perceived and the form of the 

1 Westcott and Hort, op. cit., pp. i34f. 

From Griesbach to the Present 133 

sentence or context seemed to commend abruptness; free inter- 
change of conjunctions; free interchange of the formulae introduc- 
tory to spoken words; free interchange of participle and finite verb 
with two finite verbs connected by a conjunction; substitution of 
compound verbs for simple as a rule, but conversely where the 
compound verb of the true text was difficult or unusual ; and sub- 
stitution of aorists for imperfects as a rule, but with a few examples of 
the converse. . . . 

Another impulse of scribes abundantly exemplified in Western 
readings is the fondness for assimilation. In its most obvious form 
it is merely local, abolishing diversities of diction where the same 
subject matter recurs as part of two or more neighbouring clauses 
or verses, or correcting apparent defects of symmetry. But its most 
dangerous work is 'harmonistic' corruption, that is, the partial or 
total obliteration of differences in passages otherwise more or less 
resembling each other. 1 

(3) The Alexandrian text, according to Westcott and Hort, is 
preserved to a greater or less extent in codex Ephraemi (C), 
codex Regius (L), codex 33, and the Coptic versions (especially 
the Bohairic), as well as the quotations of the Alexandrian 
Fathers, Clement, Origen, Dionysius, Didymus, and Cyril. Its 
characteristic is that which might be expected from the influence 
of a Greek literary centre — a delicate philological tact in correct- 
ing forms, syntax, and in subtle changes made in the interest of 
attaining a greater degree of polish in language and style (such 
as the rearrangement of the order of words to avoid hiatus). 

(4) The Neutral text, as its question-begging name implies, is, 
in the opinion of Westcott and Hort, the most free from later 
corruption and mixture, and comes nearest to the text of the 
autographs. It is best represented by codex Vaticanus (B), and 
next by codex Sinaiticus (X). The concurrence of these two 
manuscripts is very strong, and cannot be far from the original 
text. With the exception of a few passages, which they specify, 
Westcott and Hort declare: 

It is our belief (1) that the readings of XB should be accepted 
as the true readings until strong internal evidence is found to the 
contrary, and (2) that no readings of X B can safely be rejected 
absolutely, though it is sometimes right to place them only on an 
alternative footing, especially where they receive no support from 
Versions or Fathers. 2 

1 Ibid., pp. 122-4. 2 Ibid., p. 225. 

826158 K 


The Modern Critical Period 

The exceptions to their preference for the Neutral text are 
several passages which Westcott and Hort term 'Western non- 
interpolations'. They doubtless chose this cumbersome nomen- 
clature simply because they could not bring themselves to refer 
directly to 'Neutral interpolations' — which is exactly what, on 
their own reconstruction, is involved in these readings. In 
several passages in the last three chapters of Luke, and one in 
Matthew, 1 the Western text is regarded by Westcott and Hort 
as preserving the original form of text. The reason they aban- 
don the testimony of X and B in these passages is that here the 
Western text, which normally is the fuller and more circum- 
stantial form of text, has resisted (so they believe) the impulse to 
add material, whereas it is the Neutral text that presents the 
expanded reading. 

In accord with Westcott and Hort's critical reconstruction, 
the relation of their four text-types to the autograph may be 
represented by the following stemma : 



(D Old Lat Syr c 
fam. so far as 

X (Common ancestor of both 
Alexandrian and Neutral) 


SYRIAN ( (A) E F G H S V Y most minuscules) 


1 The Western non-interpolations, which Westcott and Hort print within double 
brackets, are in Matt, xxvii. 49; Luke xxii. 19-20; xxiv. 3, 6, 12, 36, 40, 51, and 
52. Nearly a score of other, somewhat similar passages throughout the Gospels 
form an intermediate class which, in their opinion, may also involve Western 
non-interpolations; see Westcott and Hort, op. cit., p. 176. 

From Griesbach to the Present 135 

By way of retrospect and evaluation it may be said that 
scholars today generally agree that one of the chief contribu- 
tions made by Westcott and Hort was their clear demonstration 
that the Syrian (or Byzantine) text is later than the other types 
of text. Three main types of evidence support this judgement: 

( 1 ) the Syrian text contains combined or conflate readings which 
are clearly composed of elements current in earlier forms of text; 

(2) no ante-Nicene Father quotes a distinctively Syrian reading; 
and (3) when the Syrian readings are compared with the rival 
readings their claim to be regarded as original is found gradually 
to diminish, and at last to disappear. 1 

It was perhaps not surprising that Westcott and Hort's total 
rejection of the claims of the Textus Receptus to be the original 
text of the New Testament should have been viewed with alarm 
by many churchmen. During the closing decades of the nine- 
teenth century the traditional text found a doughty defender in 
the person of John W. Burgon (1813-88), Dean of Chichester. 
He has been described as 'a High-churchman of the old school' 
who became notorious as 'a leading champion of lost causes and 
impossible beliefs; but the vehemence of his advocacy somewhat 
impaired its effect'. 2 His conservatism can be gauged from a 
sermon he preached at Oxford in 1 884 in which he denounced 
the higher education of 'young women as young men' as 'a 
thing inexpedient and immodest' ; the occasion was the admission 
of women to university examinations ! 

The publication in 1881 of the Revised Version of the King 
James or Authorized Version of 161 1 aroused Burgon's indig- 
nation not only on the score of its unidiomatic English, but even 
more because the Revisers had adopted an underlying Greek 
text substantially identical with that of Westcott and Hort. In a 
series of three learned articles in the London Quarterly Review, 
which were reprinted in a volume entitled The Revision Revised 
(London, 1883), Burgon used every rhetorical device at his dis- 
posal to attack both the English Revision and the Greek Testa- 
ment of Westcott and Hort. Burgon's argument was basically 
theological and speculative. As an ardent high-churchman he 
could not imagine that, if the words of Scripture had been 
dictated by the inspiration of the Holy Spirit, God would not 

1 Ibid., pp. 93-119. 

* So the Dictionary of National Biography, suppl. vol. i, s.n. 

136 The Modern Critical Period 

have providentially prevented them from being seriously cor- 
rupted during the course of their transmission. Consequently it 
was inconceivable to Burgon that the Textus Receptus, which 
had been used by the Church for centuries, could be in need 
of the drastic revision which Westcott and Hort had adminis- 
tered to it. (See below, p. 270.) 

What Burgon was apparently unable to comprehend was the 
force of the genealogical method, by which the later, conflated 
text is demonstrated to be secondary and corrupt. Instead of 
following the text of the few earlier manuscripts, Burgon pre- 
ferred the readings supported by the majority of the later 
witnesses. 1 Consequently, so far from sharing Westcott and 
Hort's high regard for the testimony of codex Vaticanus and 
codex Sinaiticus, Burgon maintained that, with the single 
exception of D, which exhibits the wildest text of all, the two 
manuscripts honoured by Westcott and Hort are the most 
depraved. He assures his readers 

without a particle of hesitation, that X B D are three of the most 
scandalously corrupt copies extant: — exhibit the most shamefully muti- 
lated texts which are anywhere to be met with: — have become, by 
whatever process (for their history is wholly unknown), the deposi- 
tories of the largest amount of fabricated readings , ancient blunders, and 
intentional perversions of Truth, — which are discoverable in any known 
copies of the Word of God. 2 

1 Burgon found an ally in Thomas R. Birks, Honorary Canon of Ely, in his 
Essay on the Right Estimation of Manuscript Evidence in the Text of the New Testament 
(London, 1878). Birks, who attempted to assign a mathematical weight to indi- 
vidual manuscripts, considered the later manuscripts in general to be more valuable 
than the earlier ones! 

The anachronistic views of Burgon have been resuscitated recently by Edward 
F. Hills in his booklet, The King James Version Defended! A Christian View of the New 
Testament Manuscripts (Des Moines, 1 956) , in which the author outdoes Burgon in 
defending the Textus Receptus, arguing even for the genuineness of the Comma 
Johanneum of 1 John v. 7-8. See also Hills's introduction entitled 'Dean Burgon in 
the Light of Criticism' in the 1 959 reprint of Burgon's book, The Last Twelve Verses 
of the Gospel according to St. Mark Vindicated . . . (first published in 1871). 

2 The Revision Revised (London, 1883), p. 16 (the italics are Burgon's). 

At the suggestion of Prebendary Edward Miller, Burgon's literary executor, a 
debate was- held at New College, Oxford, on 6 May 1897, when the position of 
Burgon was upheld by Miller, G. H. Gwilliam, and A. Bonus, against William 
Sanday, A. C. Headlam, and Willoughby C. Allen, who advocated the textual 
views of Westcott and Hort. One of the chief points of contention was the date of the 
Peshitta Syriac version of the New Testament. Miller maintained that this version, 
which is a witness to the Syrian type of text, goes back to the second century, and 
that therefore the Syrian type of text did not originate with Lucian and his con- 

From Griesbach to the Present 137 

Two other British scholars, F. H. A. Scrivener and George 
Salmon, were also critical of Westcott and Hort's theories, but 
were far more temperate than Burgon in the expression of their 
dissent. The former objected to Hort's total rejection of the 
testimony of the Syrian text, 1 and the latter complained that 
more consideration should have been given to the weight of 
purely Western readings. 2 

This somewhat lengthy account of the work of Westcott and 
Hort may be concluded with the observation that the over- 
whelming consensus of scholarly opinion recognizes that their 
critical edition was truly epoch-making. They presented what is 
doubtless the oldest and purest text that could be attained with 
the means of information available in their day. Though the 
discovery of additional manuscripts has required the realign- 
ment of certain groups of witnesses, 3 the general validity of their 
critical principles and procedures is widely acknowledged by 
textual scholars today. 

During his long and fruitful life Bernhard Weiss (1827-19 18), 
Professor of New Testament Exegesis at Kiel and at Berlin, 
edited the New Testament in Greek (3 vols., Leipzig, 1894- 
1900; second, small ed., 3 vols., 1902-5). Primarily an exegete, 
Weiss brought to his task an extensive and detailed knowledge 
of the theological and literary problems of the text of the New 
Testament. Instead of grouping manuscript authorities and 
evaluating variants in terms of external support, Weiss dis- 
criminated among readings in accord with what he deemed 
to be the most appropriate meaning in the context. His pro- 
cedure was to go through each New Testament book with a 
critical apparatus and to consider important textual variants, 

temporaries at the beginning of the fourth century. Sanday acknowledged that the 
.date of the Peshitta was 'the sheet anchor' of Miller's position, but was unable to 
produce convincing evidence for its later origin (The Oxford Debate on the Textual 
Criticism of the New Testament [London, 1897], p. 28). 

A few years later F. C. Burkitt set forth evidence in his monograph, St. Ephraim's 
Quotations from the Gospel (Cambridge, 1901), showing that the New Testament 
quotations in the genuine works of Ephraem (d. 373) agree largely with the Old 
Syriac version and not with the Peshitta, and that therefore the latter appears to 
have been produced subsequently to Ephraem's lifetime. 

1 F. H. A. Scrivener, A Plain Introduction to the Criticism of the New Testament, 
4th ed., ii (London, 1894), pp. 287 f. 

2 George Salmon, Some Thoughts on the Textual Criticism of the New Testament 
(London, 1897), pp. 129 ff. 

3 See pp. 214-17 below. 

138 The Modern Critical Period 

selecting in each case that reading which seemed to him 
to be justified, as Hort would have said, by intrinsic prob- 
ability. While this procedure is certainly 'subjective' to an 
extreme, one must not suppose that other methods are entirely 
'objective'. Even Westcott and Hort's criticism is subjective, for 
first they chose the method which they decided to follow, and 
then they judged that their so-called Neutral text is generally to 
be preferred above all other types of text. 

After Weiss had edited his text by adopting the variants which 
he regarded as most appropriate to the author's style and theo- 
logy, he drew up lists of different classes of error which he ob- 
served among the variant readings, and evaluated each of the 
chief Greek manuscripts in accordance with its relative free- 
dom from such faults. The classes of error which he detected are : 
(a) harmonizations among the Gospels; (b) the interchange 
of words ; (c) omissions and additions ; (d) alterations of word 
order; and (e) orthographical variation. In assessing the degree 
of freedom of Greek manuscripts from these errors, Weiss de- 
cided that codex Vaticanus was the best. It is not surprising, 
therefore, that the general complexion of Weiss's edition is 
remarkably similar to that of Westcott and Hort, who relied so 
largely on codex Vaticanus. The importance of Weiss's text is 
not only that it represents the mature opinion of a great exe- 
getical scholar who had given years of detailed consideration to 
the meaning of the text ; it is important also because the results 
of his 'subjective' methodology confirm the results of scholars 
who followed a different procedure, sometimes regarded as more 
'objective' because it started from the grouping of the manu- 
scripts themselves. 1 

Though issued in 19 10 (the text remained unchanged in the 
second edition, 1947) Alexander Souter's Greek New Testament 
represents the position of British textual scholarship as it was in 
1 88 1 ; the edition merely reproduces the Greek text that Arch- 
deacon Edwin Palmer, a member of the New Testament panel 
of British translators, constructed as the text which inferentially 
lies behind the Revised Version of 1 88 1 . Taking the third edition 
of Stephanus (1550) as the basis of his edition, Palmer formed 

1 See C. R. Gregory, 'Bernhard Weiss and the New Testament', American Journal 
of Theology,! (1896), pp. 16-37, and Kirsopp Lake, 'Dr. Weiss's Text of the Gospels: 
the Thoughts of a Textual Critic on the Text of an Exegete', ibid, vii (1903), 
PP- 249-58. 

From Griesbach to the Present 139 

a continuous text which represents the textual decisions of the 
Revisers. When, however, the English Revision was considered 
to represent correctly either of two competing readings, Palmer 
did not ordinarily alter the Textus Receptus. As a result the 
orthography, the spelling of proper names, and the typographical 
peculiarities or errors of Stephanus are, with a few exceptions, 

Souter's contribution in 19 10 was the preparation of a se- 
lected critical apparatus to go with Palmer's text. The chief 
strength of this apparatus lies in the relatively full evidence which 
is quoted from the Church Fathers, particularly the Latin 
Fathers. In 1947 the apparatus was enlarged by the addition of 
evidence from the Chester Beatty papyri and other witnesses 
brought to light since 19 10. As regards textual complexion, 
Souter's edition is closer to the Textus Receptus than is any 
other widely used Greek Testament today. 1 

The most monumental edition of the Greek New Testa- 
ment that has appeared thus far in the twentieth century is von 
Soden's Die Schriften des Neuen Testaments in ikrer dltesten erreichbaren 
Textgestalt hergestellt auf Grund ikrer Textgeschichte; I. Teil, Unter- 
suchungen (Berlin, 1902-10); II. Teil, Text mit Apparat (Gottin- 
gen, 1913). 2 Hermann Freiherr von Soden was born in 1852 at 
Cincinnati, Ohio, and died in 1914 as the result of a mishap that 
occurred while boarding a subway train in Berlin. Containing 
the results of prolonged investigation of the Greek minuscules 
and of intensive study of the history of the Greek text, von 
Soden's edition has nevertheless been described as 'a magnifi- 
cent failure'. 

Through the financial assistance of a friend, Miss Elise 
Koenigs, von Soden was enabled to send out a considerable 
number of research students and scholars to examine manu- 
scripts in the libraries of Europe and the Near East. These 

1 On the basis of an analysis of eleven sample chapters scattered through- 
out the New Testament J. Harold Greenlee found that: Nestle's text differs from 
the Textus Receptus in 233 cases, Merk's text differs in 160 cases, Bover's text in 
in cases, Vogels's text in 67 cases, and Souter's in 47 cases. In terms of percentage, 
this means that Nestle differs from the Textus Receptus 496% in comparison with 
Souter! See Kurt Aland, who quotes Greenlee's statistics in his article 'The Position 
of New Testament Textual Criticism' in Siudia Evangelica, ed. by Aland, F. L. Cross, 
et al. (= Texte und Untersuchungen, lxxiii, Berlin, 1959), p. 719. See below, p. 270. 

2 Von Soden also published a smaller edition (Handausgabe) , entitled Griechi- 
sches Neues Testament, Text mit kurzem Apparat (Gottingen, 1 91 3). 

140 The Modern Critical Period 

helpers secured partial or complete collations of an immense 
number of hitherto unexamined manuscripts. Utilizing this in- 
formation, von Soden set forth his views regarding the history 
of the text in 2,203 pages of prolegomena, much of it printed in 
small type ! Dissatisfied with the sigla previously used to desig- 
nate uncial and minuscule manuscripts, he devised a new system 
of nomenclature which indicates the age, contents, and type of 
each manuscript. Though the system is ingenious, it is also so 
highly complicated 1 that most subsequent textual critics have 
refused to adopt it, preferring instead the old system, which was 
slightly revised by Gregory 2 in order to remove a number of 
anomalies. As a result, in order to make practical use of von 
Soden's apparatus, it is necessary to consult a 'key' to unlock the 
significance of what otherwise are meaningless hieroglyphs. 3 

Von Soden's classification of the text-types of manuscripts of 
the Gospels is based on considerations of their general textual 

1 Von Soden divides all known Greek manuscripts into three classes: (1) the 
S-manuscripts, containing the whole New Testament (SiaflijKi;) , with or without the 
exception of the Book of Revelation ; (2) the f-manuscripts, containing the Gospels 
(t iayyeXiov) ; (3) the a-manuscripts, containing Acts and the Epistles, with or with- 
out the Book of Revelation (airooToXos) . Within each of these classes numbers are 
assigned in accord with the date and contents of each manuscript. The S- and a- 
manuscripts, up to the close of the ninth century, are numbered 1-49; those of the 
tenth century, 50-90 ; for the following centuries numbers of three digits are used, 
and the numeral in the hundreds place indicates the century (thus, 146 is a manu- 
script of the eleventh century ; 446, a manuscript of the fourteenth century) . In 
S-manuscripts the presence of the Book of Revelation is indicated by using 1-49 
in each hundred and 50-99 for those without it (thus S 421 would be a fourteenth- 
century manuscript containing the whole of the New Testament; S 271 would be 
a twelfth-century manuscript containing all the books of the New Testament ex- 
cept the Apocalypse) . Similarly, the contents of the a-manuscripts are indicated 
by a still more involved system of numerals. Since the t-manuscripts are so very 
numerous, the system is modified by the addition of other digits. 

Besides being intolerably complicated, von Soden's nomenclature tells more than 
one is likely to want to know — for what useful end is gained by being informed of the 
other New Testament books which a manuscript contains in addition to the book in 
which the given variant appears? Furthermore, the system permits the addition of 
only a limited number of newly found manuscripts, and makes no provision at all 
for the Greek lectionaries. Moreover, since opinion concerning the date of indi- 
vidual manuscripts may change, the elaborate provision to represent the age of a 
given document may well perpetuate what has come to be regarded as an error. 

2 Caspar Ren£ Gregory, Die griechischen Handschriften des Neuen Testaments 
(Leipzig, 1908). 

3 For convenient 'keys' to transpose von Soden's system of enumeration into the 
Gregory system, see Friedrich Kruger, Schliissel zu von Sodens Die Schriflen des Neuen 
Testaments . . . (Gottingen, 1927), and Benedikt Kraft, Die ^eicAen/ar die wichtigeren 
Handschriften des griechischen Neuen Testaments, 3te Aufl. (Freiburg i. Br., 1955). 

From Griesbach to the Present 141 

characteristics, on the form of the text of the pericope de adultera, 
and on the chapter divisions attached to them. Using these 
criteria, he divided the witnesses into three main groups, which 
he called the Koine, theHesychian, and the Jerusalem recensions. 

The K ( = Koivt}) text is divided into about seventeen sub- 
groups, of which K 1 is the oldest and best form. Produced by 
Lucian of Antioch (martyred a.d. 312), this text, with various 
subsequent alterations, became the prevailing text throughout 
the Byzantine Church (Westcott and Hort's Syrian text). 

The H ( = 'Hovxios) text, which von Soden traced to Hesy- 
chius of Egypt, is preserved today in the old uncials (B, X, C, 
L, A, and W), some minuscules (e.g. 33, 579, 892, and 1241), 
the Sahidic and Bohairic versions, and in the Alexandrian 
Fathers Athanasius, Didymus, Cyril, and others. It therefore 
includes what Westcott and Hort called the Neutral and the 
Alexandrian texts. 

The I (= 'IepoaoXvfjia) text, deriving probably from Euse- 
bius and Pamphilus of Caesarea in Palestine, is not preserved in 
substantial integrity in any outstanding manuscripts, but must 
be elicited from a number of authorities of mixed character- 
istics. The best witnesses in the Gospels are the uncials D and & 
and the minuscules 28, 372, 565, and 700, but so diverse are the 
textual phenomena that von Soden was compelled to posit 
seventeen sub-groups of witnesses which are more or less closely 
related to this text. 

According to von Soden these three recensions go back to a 
lost archetype, the I-H-K text, used by Origen, but already cor- 
rupted in the second century by Marcion, in the case of the Paul- 
ine Epistles, and by Tatian, in the case of the Gospels and Acts. 
The discovery and elimination of these corruptions brings us 
to the original text. 

Among the principles which von Soden followed in con- 
structing his text are the following: 

(1) When the readings of the three great recensions are 
certain, the reading supported by two recensions is generally 

(2) If two recensions have a reading which agrees with a 
parallel, the reading of the third which differs from the paral- 
lel is usually preferred. 

(3) The reading supported by Tatian is at once open to the 

142 The Modern Critical Period 

suspicion of departing from the original text. Only in the event 
of two recensions agreeing with Tatian and the dissenting 
recension agreeing with a parallel is the latter to be adjudged 
secondary; and this remains the case even when the former 
reading also agrees with a parallel. 

(4) When early, certainly mutually independent witnesses — 
even though they may be only patristic writers or versions — 
agree in a reading which differs from Tatian, this reading 
requires serious consideration for adoption even when all three 
recensions agree with Tatian. 

While acknowledging the enormous amount of research which 
von Soden's edition represents, most scholars have criticized 
his methods and results in the following respects: 1 

(a) Since von Soden tends to give preference to readings 
supported by two of the three main texts, by this procedure the 
Koine type of text is elevated to a rank co-ordinate in impor- 
tance with the other two texts. So far from regarding the Koine 
as an independent entity, however, most scholars today follow 
the view of Griesbach, Hort, and others, that this text is largely 
secondary and derivative from the others. As a consequence of 
von Soden's high estimate of the value of the Koine text, his 
edition approaches the Textus Receptus more closely than does 
any previous modern critical text. 

(b) Though von Soden thought that his chief contribution to 
textual studies was the isolation and subdivision of his /-text, 
later scholars regard it as the least sound, for he includes in one 
text-type such heterogeneous elements as the Western witnesses, 
the Caesarean text, the Old Latin, and the Old Syriac, as well 
as witnesses which are mixed with the Koine text. 

(c) While Marcion and Tatian undoubtedly had a certain 
corrupting influence upon the transmission of the New Testa- 
ment text, von Soden assigned them an altogether dispropor- 
tionate degree of importance in the contamination not only of 
the Latin and Syriac versions but of Greek witnesses as well. 

(d) Though absolute accuracy in an extensive critical appara- 
tus is probably unattainable, where von Soden's work can be 

1 See, for example, K. Lake in Review of Theology and Philosophy, iv (1908), 
pp. 201-17, 277-95; H. C. Hoskier in Journal of Theological Studies, xv (1914), 
pp. 307-26; Hans Lietzmann in £eitschrift fur die neutestamentliche Wissenschaft, 
xv (1914), pp. 323-31 ; and A. Souter in Expositor, 8th series, x (1915), pp. 429-44. 

From Griesbach to the Present 143 

tested it has been found to contain a higher percentage of errors 
than is usually considered to be consistent with trustworthy 

Despite these and other justifiable criticisms which have been 
levelled against von Soden, his edition remains a monument of 
broad research and immense industry which, with the extensive 
prolegomena dealing with the history of the transmission of the 
text, must be taken into account by every serious textual critic. 
The next three editions to be mentioned are the products 
of twentieth-century Roman Catholic scholarship. The edition 
prepared by Heinrich Joseph Vogels (Diisseldorf, 1920; with the 
Latin Vulgate, 1922; 4th ed., Freiburg i. Br., 1955) is closer to 
the Textus Receptus than the other two. 1 The editor provides a 
limited apparatus which, in addition to citing the principal 
uncial and minuscule manuscripts, is relatively full as regards 
evidence derived from the Old Latin materials and the Syriac 

Relying on von Soden's apparatus, but transposing von 
Soden's sigla to those of the Gregory system, and adding fresh 
manuscript evidence, Augustin Merk, S.J., published an edition 
of the Greek and Latin New Testament through the Pontifi- 
cal Biblical Institute (Rome, 1933; 9th ed., 1964, with an Ap- 
pendix of variant readings from recently discovered papyri). 2 
The apparatus, which includes evidence from the several Tatianic 
witnesses, is drawn up so as to show family relationship among 
the witnesses. Unfortunately, however, Merk's citation of evi- 
dence is far from accurate, 3 and when his apparatus supplies 
evidence not available for verification in other publications, one 
hesitates to rely upon his testimony. In the construction of his 
Greek text, Merk departs farther from the Textus Receptus than 
do the othertwo Roman Catholic editors. 

Jose Maria Bover, S.J., devoted his efforts over many years to 
the collection and evaluation of textual materials. 4 The Greek 

1 See footnote I on p. 139 above. 

2 Merk's text was reprinted by Gianfranco Nolli in his Novum Testamentum graece 
et latine (Rome, 1955). The footnotes of Nolli's edition supply the student with a 
limited number of text-critical, syntactical, and lexical aids. 

3 See, for example, H. G. Opitz's review in Gnomon, xii (1936), pp. 429-36, and 
G. D. Kilpatrick, 'Three Recent Editions of the Greek New Testament', Journal of 
Theological Studies, 1 (1949), pp. 145 ff. 

4 For a summary of Bover's textual investigations, see B . M. Metzger, Chapters in the 
History of New TestamentTextual Critirism(Leiden and Grand Rapids, 1 963) , pp. 1 2 1 -4 1 . 

144 The Modern Critical Period 

text of his bilingual edition (Madrid, 1943; 4th ed., 1959), 
which is printed with the beautiful font of Greek type belonging 
to the Association Guillaume Bude, is an eclectic one, departing 
frequently from the Alexandrian type of text and approaching 
the Western or Caesarean type. The apparatus, which presents 
information concerning the textual opinions of six modern edi- 
tors, supplies manuscript evidence for only the more important 

The most widely used pocket edition of the Greek Testament 
is that prepared by Eberhard Nestle (1 851-19 13) for the Wiirt- 
tembergische Bibelanstalt (Stuttgart, 1898; 24th ed., by Erwin 
Nestle and Kurt Aland, i960). Its text (since the 3rd ed., 1901) 
is based on a comparison of the texts edited by Tischendorf 
(1869-72), by Westcott and Hort (1881), and by Bernhard 
Weiss (1894- 1 900) ; where two of these three editions agree, this 
reading is printed by Nestle. 1 Thus the text of Nestle represents 
the state of nineteenth-century scholarship ; its apparatus, how- 
ever, which is a marvel of condensation, supplies with a high 
degree of accuracy a great amount of textual information, in- 
cluding many early witnesses that were discovered during the 
twentieth century. 

In connexion with the sesquicentennial celebration of the 
British and Foreign Bible Society (1 804-1 954), a new edition of 
Eberhard Nestle's 1904 text was edited with an apparatus pre- 
pared by G. D. Kilpatrick with the help of Erwin Nestle and 
several other scholars (London, 1958). The text of the 1904 edi- 
tion was changed in about twenty passages (of which eleven are 
listed in the introduction) and several alterations were made 
in orthography, accentuation, and the use of brackets. As regards 
the apparatus, the number of variants cited is substantially 
smaller than those in current editions of Nestle, but a certain 
amount of additional information is provided for the variants 
that are cited. (For comments on Kilpatrick's Greek text of the 
Diglot New Testament, privately circulated by the British and 
Foreign Bible Society, see pp. 177 f. below.) 

In conclusion, reference should be made to the apparatus 

1 Since the 17th ed., however, a small number of variants which, according to 
widespread scholarly opinion, have strong claims to be original have been taken 
into the text against the majority of the three nineteenth-century editions. 

From Griesbach to the Present 145 

criticus published by S. C. E. Legg at Oxford in 1935 and 1940. 
Choosing the Greek text of Westcott and Hort as the collating 
base, Legg supplied for Mark (1935) and for Matthew (1940) 
an enormous thesaurus of variant readings of Greek manuscripts, 
early versions, and patristic quotations. It is regrettable that 
Legg did not indicate in every case the editions of the versions 
and Fathers on which he relied. He has been criticized also for 
incomplete citation of evidence as well as occasional errors. 1 
Despite such justifiable criticisms of faults which arise chiefly 
from the ambitious scope of Legg's project — a project which 
probably exceeded the capacity of any single scholar to ac- 
complish — these two volumes present an extraordinary amount 
of textual information, surpassing any previous apparatus for 
Matthew and Mark. 2 

Shortly after the publication of the New Testament of the 
New English Bible (1961), requests were received by the Oxford 
and Cambridge University Presses to issue an edition of the 
Greek text that inferentially lies behind the new English version. 
R. V. G. Tasker, a member of the Panel of Translators of the 
N.E.B., was entrusted with the task of preparing the edition, 
which was published in 1964. In an Appendix Tasker cites 
manuscript evidence 3 for about 270 sets of variant readings that 
are represented in the margin of the N.E.B. 4 

In 1966, after a decade of work by an international Com- 

1 See, for example, E. C. Colwell, Classical Philology, xxxiii (1938), pp. 112-15; 
A. Souter, Expository Times, liii (1941), pp. 169 ff.; G. D. Kilpatrick, Journal of 
Theological Studies, xliii (194,2), pp. 30-34; and T. W. Manson, ibid., pp. 83-92. 

2 In 1949 an international project was launched in order to pool the resources of 
textual scholars in Britain and America for the production of a comprehensive 
apparatus criticus of the Greek New Testament. Work is in progress on the Gospel of 
Luke. For a description of the aims of the project, see M. M. Parvis in Crozer. 
Quarterly, xxvii (1950), pp. 301-8; for a report of progress, see an article by E. C. 
Colwell et al. to appear in Journal of Biblical Literature, lxxxvii ( 1 968) . 

3 The very important Bodmer Papyri, however, are conspicuous by their absence, 
for most of them were published after the N.E.B. Panel had completed its work. 

4 For an evaluation of Tasker' s edition, reference may be made to T. Gaumer, 
'An Examination of Some Western Textual Variants Adopted in the Greek Text 
of the New English Bible', The Bible Translator, xvi (1965), pp. 184-9, an d to 
the present writer's volume, Historical and Literary Studies, Pagan, Jewish, and Christian 
(Leiden and Grand Rapids, 1968), pp. 160 f. 

146 The Modern Critical Period 

mittee, 1 five Bible Societies 2 published an edition of the Greek 
New Testament designed for the use of Bible translators and 
students. The textual apparatus, which provides a relatively full 
citation of manuscript evidence, includes about 1440 sets of 
variant readings, chosen especially in view of their exegetical 
significance. There is also a punctuation apparatus that cites 
meaningful differences of punctuation in about 600 passages, 
drawn from five editions of the Greek New Testament and from 
ten translations in English, French, and German. A companion 
volume, providing a summary of the Committee's reasons for 
adopting one or another of the variant readings, has been pre- 
pared by the present writer, and is scheduled for publication 
toward the close of 1 968. 

The preceding survey of the more important printed editions 
of the Greek New Testament has referred to only a relatively 
small proportion of the total number of editions. No one 
knows exactly how many separate editions of the Greek Testa- 
ment have come from the press since 1514, but it is undoubtedly 
a very great number indeed. Eduard Reuss of Strasbourg, who 
published a description of editions issued up to 1869, was able to 
enumerate 584 separate editions. 3 If one adds the re-editions, the 
variant editions, and some doubtful editions which Reuss men- 
tions in part, the number amounts to 853. Furthermore, since 
Reuss' s list is not complete for the more recent period which 
he covers, and since many editions have appeared since he pub- 
lished his volume, it is altogether probable that the 1,000 mark 
was passed early in the twentieth century. 

» The members of the Committee are Kurt Aland of Munster, Matthew Black 
of St. Andrews, Allen Wikgren of Chicago, and the present writer. During the 
first four years of its work the Committee also included the Estonian scholar 
Arthur Voobus. For the preparation of the second edition, scheduled to appear 
in 1968, the Committee was enlarged by the addition of Carlo M. Martini, S.J., 
of Rome. 

2 They are the American, British and Foreign, Dutch, Scottish, and Wilrttem- 
berg Societies. In 1967 the edition was published under the auspices of the United 
Bible Societies. 

3 Eduard Reuss, Bibliotheca Novi Testamenti graeci cuius editiones ab initio typo- 
graphiae ad nostram aetatem impressas quoquot reperiri potueranl (Brunsvigae, 1872). 
For additions to Reuss's list, see Isaac H. Hall in Philip Schaff, A Companion to 
the Greek Testament and the English Version, 3rd ed. (New York, 1889), pp. 519-24. 


The Application of Textual Criticism 
to the Text of the New Testament 


The Origins of Textual Criticism 
as a Scholarly Discipline 

II ke so many disciplines that we take for granted in our 
Western culture, textual criticism originated among the 
JL Greeks. Its rise and development were connected with the 
Homeric epics. Because the rhapsodists who recited portions of 
the Iliad and the Odyssey in public would occasionally alter the 
text to suit the special occasion or their own notion of an effec- 
tive arrangement, there were many versions current even in 
very early times. There subsequently grew up several 'City 
Editions' of Homer, namely those which presumably were pre- 
served by civic authority in various centres (traditionally, seven 
centres) and from which private copies were made. Other spe- 
cial texts were made by Theagenes of Regium, by Stesimbrotus 
of Thasos (c. 450 B.C.), and by Aristotle, who prepared a version 
for his pupil Alexander the Great, usually called 17 eV rov vdpd-qxos 
from the case in which it was kept (Plutarch, Alex. 8) . 

A more scientific criticism of the text of Homer was developed 
in the Hellenistic Age. This critical study was pursued at the 
famed library in Alexandria, which was reputed to have about 
600,000 volumes, 1 and for which, according to tradition, the 
Greek translation of the Old Testament, the so-called Septua- 
gint, was made. The early directors of the library sought to pro- 
vide ever more accurate editions of the Homeric poems. Shortly 
before 274 b.c. the first of these scholarly librarians, Zenodotus 
of Ephesus (c. 325-c. 234B.C.) , made a comparison of many manu- 
scripts in order to restore the original text of both the Iliad and 
the Odyssey. The corrections which Zenodotus made in the text 
of Homer were of four kinds : ( 1 ) he eliminated verses which he 
regarded as spurious; (2) he marked others as doubtful but left 
them in his edition; (3) he transposed the order of verses; and 
(4) he introduced new readings not generally current. 

1 See Edward A. Parsons, The Alexandrian Library, Glory of the Hellenic World: Its 
Rise, Antiquities, and Destructions (Amsterdam, London, and New York, 1952). 

826158 L 

150 The Origins of Textual Criticism as a 

One of the subsequent directors of the library was Aristo- 
phanes of Byzantium (c. 257-c. 180 B.C.), perhaps the most 
distinguished philologist of Greek antiquity, to whom is as- 
cribed the invention of the Greek accent marks as well as other 
diacritical signs. In his edition of the Iliad and the Odyssey 
Aristophanes employed a variety of critical symbols to indicate 
his opinion of the state of the text thus marked. His greatest 
pupil was Aristarchus of Samothrace (c. 220-t. 144 B.C.), who, 
becoming his successor at the library, edited the works of half 
a dozen Greek authors and published two critical editions of the 
Homeric poems, supplementing the number of critical symbols 
which his predecessors had used. 

Thus there was a fairly well-developed scholarly discipline of 
textual and literary criticism in antiquity, localized chiefly at 
Alexandria and directed primarily toward the epics of Homer. 
It is common knowledge that Philo Judaeus and many Church 
Fathers, influenced by the philological scholarship current at 
Alexandria, utilized in their interpretation of the Scriptures the 
methods of allegorical exegesis which had been applied to cer- 
tain stories of the gods and goddesses included in the Homeric 
cycle. It is less widely appreciated — indeed, the question has 
seldom been raised — how far the methods of textual criticism 
current at Alexandria were adopted by scholars in the Church 
and applied to the text of the New Testament. The following is a 
brief summary of what can be learned from patristic sources 
relating to this subject. 

Ironically enough, the earliest efforts 1 to ascertain the original 
text of the New Testament seem to have been made by those who 
were excommunicated as heretics by the authoritarian Bishop 
of Rome, Pope Victor (a.d. 187-98). It appears that a learned 
leather-merchant (okvtcvs) named Theodotus, lately come from 
Byzantium to Rome, had been stung by certain criticisms which 
Galen, the famous Greek physician, had levelled against the 
philosophical naivete of many Christians. 2 In an attempt to 
introduce improvements in the methodology of scriptural inter- 
pretation, Theodotus and his followers seem to have undertaken 
a critical recension of the Biblical text. Eusebius preserves a 

1 The alterations made by Marcion in the New Testament were motivated by 
doctrinal considerations rather than by an interest in textual criticism. 

2 See R. Walzer, Galen on Jews and Christians (Oxford, 1949), pp. 75 ff. 

Scholarly Discipline 151 

large excerpt of an almost contemporary pamphlet by an anony- 
mous author directed against these philosophically minded 
Christians. 1 According to this author the Theodotians deserved 
to be condemned on three scores : ( 1 ) they were engrossed in the 
study of logic, mathematics, and empirical science ('Some of 
them, in fact, study the geometry of Euclid, and admire Aristotle 
and Theophrastus ; and Galen perhaps is even worshipped by 
some of them') ; (2) abjuring allegorizing, they practised strict 
grammatical exegesis ; and (3) they applied textual criticism to 
the Septuagint and the Greek New Testament ('They did not 
fear to lay hands on the divine Scriptures, alleging that they had 
critically revised (SicjpOcjKevai) them. . . . For they cannot deny 
that this audacious act is their own, seeing that the copies are 
written in their own hand, and they did not receive the Scrip- 
tures in this condition from their teachers, nor can they show any 
copies from which they made their emendations (Seigai avTiypa<f>a 
odev aura jxeTeypdipavTO jxr) e)(W(nv)'). z Unfortunately, nothing 
more is known of this early effort at textual criticism. 

Not long after the Theodotians had been excommunicated, 
one of the most assiduous and erudite scholars of his age, 
Origen of Alexandria and Caesarea, began a text-critical study 
of the entire Old Testament in Hebrew and in several Greek 
translations. His resulting Hexapla, which must have required 
many years of the most painstaking labour, was a monumental 
tool that many patristic scholars consulted, in the famed library 
of Pamphilus at Caesarea, until its destruction in the seventh 
century during the Islamic conquest of the Near East. 

The question whether Origen ever attempted to edit a critical 
text of the New Testament has been answered quite diversely by 
modern scholars; 3 it seems probable to the present writer that 
he did not extend his textual efforts to preparing a formal 
edition of the New Testament. At the same time, in all his 

1 Eusebius, Hist. Eccl. v. xxviii. 13-19 (the excerpt may be from Hippolytus of 
Rome's Little Labyrinth). 

1 For a discussion of this passage, see Hermann Schone, 'Ein Einbruch der anti- 
ken Logik und Textkritik in die altchristliche Theologie', in Pisciculi: Studien zur 
Religion und Kultur des Altertums; Franz Joseph Dolger . . . dargeboten . . . (Mfinster in 
W., 1939), pp. 252-65. 

3 For a summary of these opinions, reference may be made to B. M. Metzger, 
'Explicit References in the Works of Origen to Variant Readings in New Testament 
Manuscripts', in Biblical and Patristic Studies in Memory of Robert Pierce Casey, ed. by 
J. N. Birdsall and R. W. Thomson (Freiburg, 1963), pp. 78-95. 

152 The Origins of Textual Criticism as a 

writings and particularly in his exegetical treatises, Origen re- 
veals a certain solicitude for critical details in the Biblical text. 
He complains that 'the differences among the manuscripts [of 
the Gospels] have become great, either through the negligence of 
some copyists or through the perverse audacity of others ; they 
either neglect to check over what they have transcribed, or, 
in the process of checking, they lengthen or shorten, as they 
please'. 1 Besides making comments of a general nature about the 
text, Origen sought out information (though he did not always 
utilize this information) concerning variant readings in Greek 
manuscripts of the New Testament. He observes, for example, 
that in Matthew's account (xviii. i) of the disciples' question 
as to who is the greatest in the kingdom of heaven, according 
to some of the manuscripts the Evangelist prefixed the phrase 
iv eK€ivr\ rfj u>pa, whereas according to others the expression 
iv eKeivT) rfj 17/xepa appears. 2 Similarly, Origen notices the two 
readings in Heb. ii. 9, 'apart from God' {x<*>pls d*ov) and 'by the 
grace of God' {xdpiTi deov), but is not interested in deciding be- 
tween them, for he finds spiritual significance in both readings. 
At other times Origen declared his preference among variant 
readings, but often his choice appears to be based on considera- 
tions other than those of a purely textual nature. Thus when 
he dismisses the reading 'Jesus Barabbas' in favour of simply 
'Barabbas' (Matt, xxvii. 16-17), ne does so because he thinks 
that the name 'Jesus' was never applied to evil-doers. 3 Again, 
Origen's well-known preference for the reading 'Bethabara' in- 
stead of 'Bethany' as the place of John's baptizing (John i. 28) 
was adopted on geographical and etymological grounds, 4 and 
the same reasons dictated his preference for 'Gergesa' rather 
than 'Gerasa' or 'Gadara' as the name of the place where the 
demons entered the herd of swine. 5 In a different category are 
instances where, because of some exegetical difficulty, Origen 
suggests that perhaps all of the manuscripts existing in his day 
may have become corrupt. 6 

1 Comm. in Matt. xv. 14 (Die grieckischen ckristlichen Schriftsteller, Origenes, x. 387. 
28-388. 7, ed. Klostermann). 

2 Comm. in Matt. xiii. 14 (G.C.S., Origenes, x. 213. 21 ff., ed. Klostermann). 

3 In Matt. Comm. ser. 121 (G.C.S., Origenes, xi. 2. 255, 24 ff., ed. Klostermann). 
+ Comm. in Joan. vi. 40 (24) (G.C.S., Origenes, iv. 149. 12 ff., ed. Preuschen). 

5 Comm. in Joan. vi. 41 (24) (G.C.S., Origenes, iv. 150. 3 ff., ed. Preuschen). 

6 For examples, see the study mentioned in n. 3 on the preceding page. 

Scholarly Discipline 153 

Judged according to modern standards, St. Jerome (c. 347- 
420) was a more sagacious textual critic than Origen, well 
aware of the varieties of errors which arise in the transcription 
of manuscripts. He refers, for example, to the possibility of con- 
fusion of similar letters, confusion of abbreviations, accidents 
involving dittography and haplography, the metathesis of let- 
ters, assimilation, transpositions, and deliberate emendations by 
scribes. 1 Several explicit references will indicate his interest in 
text-critical details. In the preface to his revision of the Latin 
Gospels, addressed to Pope Damasus, who had requested that 
he undertake the work, Jerome declares that for the textual basis 
of the revision he has relied upon older Greek manuscripts. 
Again, in his letter to Minervius and Alexander, 2 two monks at 
Toulouse who had written to Jerome asking him to explain cer- 
tain passages in Scripture, Jerome discusses several forms of the 
text of 1 Cor. xv. 51 ('We shall not all sleep, but we shall all be 
changed'). He indicates that he prefers the reading, 'We shall all 
sleep, but we shall not all be changed'. In his Dialogue against 
the Pelagians, 3 Jerome states that in certain copies, and espe- 
cially in Greek codices, an extensive addition was to be found 
at the close of the Gospel according to Mark. Jerome does not 
tell us where he found these manuscripts, and no such copy 
was known until the twentieth century, when the passage turned 
up in a Greek manuscript which Mr. Charles L. Freer of De- 
troit had bought from an Arab dealer in Gizeh near Cairo (for 
the translation of this addition, see p. 57 above). 

Though primarily a theologian, St. Augustine (354-430) 
showed on occasion a keen critical judgement in textual prob- 
lems. Thus when considering the difficulty that Matthew 
(xxvii. 9) attributes a quotation to Jeremiah which actually 
appears in Zechariah, Augustine suggests that one should 'first 
take notice of the fact that this ascription of the passage to 
Jeremiah is not contained in all the manuscripts of the Gospels, 
and that some of them state simply that it was spoken "by the 
prophet". It is possible, therefore, to affirm that those manu- 
scripts deserve rather to be followed which do not contain the 

1 For one or more examples from the works of Jerome illustrating each of these 
categories, see K. K. Hulley, 'Principles of Textual Criticism Known to St. Jerome', 
Harvard Studies in Classical Philology, lv (1944), pp. 87-109. 

2 Epist. 119 (Migne, Patrologia Latina, xxii. 966 ff.). 

3 Dialog, contra Pelagianos, ii. 15 (Migne, P.L. xxiii. 576). 

154 The Origins of Textual Criticism as a 

name of Jeremiah. For these words were certainly spoken by 
a prophet, only that prophet was Zechariah. . . .' With com- 
mendable candour, however, Augustine declares that he is not 
altogether satisfied with this explanation, because 'a majority 
of manuscripts contain the name of Jeremiah, and those who 
have studied the Gospel with more than usual care in the Greek 
copies report that they have found it to stand so in the more 
ancient Greek exemplars'. Thereupon Augustine virtually enun- 
ciates the critical canon that the more difficult reading is to 
be preferred; he continues, T look also to this further considera- 
tion, namely that there was no reason why this name should 
have been added [subsequently to the true text] and a corrup- 
tion thus created; whereas there was certainly an intelligible 
reason for erasing the name from so many of the manuscripts. 
For presumptuous inexperience (audax imperitia) might readily 
have done that, when perplexed with the problem presented by 
the circumstance that this passage cannot be found in Jeremiah.' 1 

On another occasion Augustine suggests that preference 
should be given to readings that are current in important sees, 
thus anticipating B. H. Streeter's theory of 'local texts' (see 
pp. 169-73 below) . He writes : 'If the books of the New Testament 
are confusing in the variety of their Latin translations, they should 
certainly give place to the Greek versions, especially to those 
which are found among the more learned and diligent churches.' 2 

During the Middle Ages, when a knowledge of Greek was 
at a low ebb, text-critical efforts were now and then directed 
toward the purification of Jerome's Vulgate text. It was per- 
haps to be expected that this version, besides being corrupted 
with the usual types of error incident to all transcription, would 
once again incorporate certain Old Latin readings which Jerome 
had eliminated from his text. (For some of these attempts to 
purify Jerome's text, see p. 76 above.) The writings of such 
authors as Gilbert of Porree and Peter Lombard contain spo- 
radic comments, reflecting information derived from Jerome 
and Augustine, regarding the Greek lying behind such and such 
Latin renderings. 3 

1 De consensu Evangel, iii. 7. 29 (Migne, P.L. xxxiv. 11741".). 

1 De doctr. Christ, ii. 15. 22 (Migne, P.L. xxxiv. 46), 'apud ecclesias doctiores et 

3 See Arthur Landgraf, 'Zur Methode der biblischen Textkritik im 12. Jahr- 
hundert', Biblica, x (1929), pp. 445-74. See below, p. 270. 

Scholarly Discipline 155 

At the time of the Renaissance and with the spread of the 
knowledge of ancient Greek, scholars began to correct the Latin 
Vulgate by the original Greek. In their Biblical annotations 
Erasmus and Beza not infrequently refer to variant readings in 
Greek manuscripts. As was mentioned in Chapter III, the first 
English Bible to contain the translation of variant readings from 
Greek manuscripts (including codex Bezae) was the Geneva 
Bible of 1560, prepared by William Whittingham and other 
English exiles residing at Geneva. This Bible has, for example, 
the negative Golden Rule in the margin opposite Acts xv. 29, 
as well as a literal translation of the Western variant at Acts 
xix. 9, that Paul preached daily at the school of Tyrannus 'from 
five a clocke unto ten'. 1 

The first scholar to make any use of all three classes of evi- 
dence for the text of the New Testament — that is, Greek manu- 
scripts, the early versions, and quotations from the Fathers — 
was probably Francis Lucas of Bruges (Brugensis) in his Nota- 
tiones in sacra Biblia, quibus variantia . . . discutiuntur (Antwerp, 
1 580) . Toward the close of the seventeenth century the scientific 
foundations of New Testament criticism were laid in four monu- 
mental publications of Richard Simon (1638-1712), a French 
Catholic scholar far ahead of his day in Biblical research. 
The volumes are entitled Histoire critique du texte du Nouveau 
Testament (Rotterdam, 1689); Eng. trans., Critical History of 
the Text of the New Testament, 2 parts (London, 1689); Histoire 
critique des versions du Nouveau Testament (Rotterdam, 1690); 
Eng. trans., Critical History of the Versions of the New Testament 
(London, 1692); Histoire critique des principaux commentateurs du 
Nouveau Testament, 2 parts (Rotterdam, 1693); and Nouvelles 
observations sur le texte et les versions du Nouveau Testament (Paris, 
1695). Disregarding the traditional and dogmatic presupposi- 
tions of his age, Simon examined critically the text of the Bible 
as a piece of literature. His works are full of acute observation 
and reasoning, and anticipate in detail many of the conclusions 
of scholars two and three centuries later. 

1 The Genevan translation of this verse is a literal rendering of the Greek ; the 
equivalent according to the modern reckoning of time is 'from eleven o'clock to 
four'. For a score of other variant readings given in the margins of the Geneva 
Bible, see the article mentioned in n. 2 on p. 105 above. 


Modern Methods of Textual Criticism 


The method of textual criticism which has been generally 
practised by editors of classical Greek and Latin texts in- 
volves two main processes, recension and emendation. 
Recension is the selection, after examination of all available 
material, of the most trustworthy evidence on which to base a 
text. Emendation is the attempt to eliminate the errors which are 
found even in the best manuscripts. 1 

The classical method of textual criticism arose during and 
after the Renaissance when attention was drawn to spurious 
papal decretals and when questions were raised regarding fal- 
sifications in Church history and in the credentials of certain 
religious orders. The critical acumen of scholars was sharpened 
likewise by the large number of forged texts that began to ap- 
pear; for example, a single forger, Giovanni Nanni (alias Joannes 
Annius, 1432-1502), a Dominican monk of Viterbo, put forth 
seventeen spurious treatises attributed by him to ancient Greek 
and Latin authors. 2 

A more critical spirit in dealing with ecclesiastical documents 
found expression during the sixteenth century in the work of 
Matthias Flacius and the group of Lutheran scholars known as 

1 For a concise account of these two basic processes, see Paul Maas, Textual 
Criticism (Oxford, 1958). Other useful discussions of the classical procedures of 
textual criticism include the following : R. C. Jebb inA Companion to Greek Studies, ed. 
by Leonard Whibley (Cambridge, 1 go6) , pp. 610-23 ; J. P. Postgate in A Companion to 
Latin Studies, ed. byj. E. Sandys (Cambridge, 1910), pp. 791-805 ; L. Havet, Manuel 
de critique verbale appliguie aux textes latins (Paris, 191 1) ; F. W. Hall, A Companion to 
Classical Texts (Oxford, 1913), pp. 108-98; H. Kantorowicz, Einfiihrung in die 
Textkritik (Leipzig, 1921) ; andL.Bieler, 'The Grammarian's Craft', in Folia: Studies 
in the Christian Perpetuation of the Classics, x (1956), pp. 3-42. See below, p. 271. 

2 Annius' volume, De Commentariis antiquitatum (Rome, 1498), includes the 
allegedly rediscovered writings of Xenophon, Berosus, Cato, Antoninus Pius, 
Manetho, Philo Judaeus, Caius Sempronius, and Myrsilus, all of which, Annius 
declared, he had found buried in the ground. See 'Nanni' in Pierre Bayle's Dictio- 
naire historique el critique, and 'Annius' in J. S. Ersch and J. G. Gruber's Allgemeine 
Encyclopddie; cf. also James Farrer, Literary Forgeries (London, 1907), pp. 67-81. 

Modern Methods of Textual Criticism 157 

the Magdeburg Centuriators, who were the first to write the 
history of the Church from a Protestant point of view. In 1675 
the Jesuit scholar Daniel Papebroch aroused the hostility of the 
Benedictines by denying the authenticity of documents con- 
stituting the credentials of certain Benedictine monasteries. The 
learned Benedictine monks at St. Maur took up the challenge 
by founding the science of palaeography, which is the classifica- 
tion of manuscripts according to their age in the light of their 
handwriting and other indications. The first treatise to deal with 
the Latin palaeography of official documents was the monu- 
mental work of the Maurist Jean Mabillon (1 632-1 707) en- 
titled De re diplomatica (Paris, 1681). The science was extended 
to Greek manuscripts by another Benedictine, Bernard de 
Montfaucon (1 655-1 741), in his Palaeographica graeca (Paris, 

The application of critical methods in the editing of classical 
texts was developed principally by three German scholars, 
Friedrich Wolf (1 759-1824), one of the founders of classical 
philology, Immanuel Bekker (1785-1871), and Karl Lachmann 
( 1793-185 1). Bekker devoted his long life to the preparation of 
critical editions of Greek texts. The transfer of many manu- 
scripts to public libraries as a result of the upheaval following 
the French Revolution gave opportunity for extensive collation 
of manuscripts older than those which had previously been 
generally available. Bekker collated some 400 manuscripts, 
grouped existing manuscripts of an author into families where 
one was derived from another, and published sixty volumes of 
improved editions of Greek authors. As was mentioned earlier 
(p. 124 above), Lachmann went further than Bekker, showing 
how, by comparison of manuscripts, it is possible to draw infer- 
ences as to their lost ancestors or archetypes, their condition, and 
even their pagination. 

The basic principle which underlies the process of con- 
structing a stemma, or family tree, of manuscripts is that, apart 
from accident, identity of reading implies identity of origin. By way of 
example, suppose that there are seven manuscripts of an ancient 
book, and that in a certain paragraph three of them agree in 
lacking a sentence which is present in the other four manuscripts. 
From this circumstance we would deduce either that a common 
ancestor of the three had omitted the sentence, or that an 

i 5 8 

Modern Methods of Textual Criticism 

ancestor of the four had added it. Suppose, moreover, that we 
find that the seven manuscripts frequently range themselves so 
that one of them (which we may designate A) stands apart, 
showing no great similarity to any of the other six, while B, 
C, and D, on the one hand, and E, F, and G, on the other hand, 
greatly resemble each other, though differing somewhat from 
the rest. We can express this by saying that B, C, and D form 
a family, descended from a hypothetical common ancestor 
which we may call X, and that E, F, and G form another family, 
descended from a hypothetical ancestor which we may call T. 
The readings of X which can be deduced by comparing those of 
BCD will be of a higher antiquity and of greater authority than 
any of the readings in B or C or D taken singly. And the same 
may be said for the readings of T when compared with those in 
E or F or G. Indeed, it is possible to go further : we may compare 
the readings of X and T with each other, and with those of A, 
and thus deduce the readings of a still more remote ancestor 
which we may call Z> the hypothetical archetype of all the 
manuscripts. Thus the pedigree of all ten manuscripts (the 
seven extant and the three hypothetical) would be as follows : 







It follows that because B, C, and D may agree in a given 
reading against A, such a reading is not three times more likely 
to be correct than the reading in A. In fact it is obvious that, 
other things being equal, there is a fifty-fifty chance that either of 
the two readings may be correct, for where B, C, and D agree 
they represent a manuscript (X) which is as far removed from 
the archetype (£) as is A. Thus, instead of merely counting the 
number of manuscripts supporting a given reading, the editor 
must weigh their significance in accord with their mutual rela- 
tions to one another. 

Often, however, difficulties hinder the construction of a 

Modern Methods of Textual Criticism 159 

stemma of manuscripts. The simple example given above as- 
sumes that the different lines of descent have remained inde- 
pendent of one another. But a disturbing element enters when 
mixture has occurred, that is, when a copyist has had two or 
more manuscripts before him and has followed sometimes one, 
sometimes the other; or, as sometimes happened, when a scribe 
copied a manuscript from one exemplar and corrected it against 
another. To the extent that manuscripts have a 'mixed' ancestry, 
the genealogical relations among them become progressively 
more complex and obscure to the investigator. 1 


During the twentieth century the genealogical method has 
come under attack from several quarters; some scholars have 
rejected it entirely, while others have restricted its application 
to a narrow and limited area. One of the former is Joseph Bedier, 
an editor of several medieval French documents. It was while 
preparing his edition of Le Lai de V Ombre par Jean Renart (Paris, 
19 1 3), so Bedier declared, that he became distrustful of the 
genealogical method, ( 1 ) because in practice it has almost always 
resulted in the construction of a tree with two branches of 
witnesses (a circumstance which Bedier viewed with cynical 
suspicion, suggesting that editors have deliberately forced the 
evidence into a stereotyped pattern), and (2) because one can 
often argue well for several different stemmata of classification 
of manuscripts. His own method was to choose what seems to be 
the best manuscript, making the choice on the basis of grammar, 
coherent sense, simple and regular orthography, and then to use 
the other manuscripts eclectically in correcting sporadic read- 
ings in the manuscript accepted as primary. 2 

1 Genealogical mixture is widespread in the textual transmission of most Latin 
patristic literature; see the sane and balanced account in Maurice Bevenot, S.J., 
The Tradition of Manuscripts: a Study in the Transmission of St. Cyprian's Treatises 
(Oxford, 1961). For the special problems which confront the editor of Greek 
patristic documents, see Herbert'Musurillo's discussion of 'Some Textual Problems 
in the Editing of the Greek Fathers', in Texte und Unlersuchungen, lxxviii (1961), 
pp. 85-96. 

2 Besides the preface in his 191 3 edition, see B^dier's fuller exposition in 'La 
Tradition manuscrite du Lai de V Ombre: Reflexions sur l'art d'editer les anciens 
textes', Romania, liv (1928), pp. 161-96, 321-56. 

160 Modern Methods of Textual Criticism 

Among New Testament scholars who have come under the 
influence of Bedier's scepticism of the value of the genealogical 
method are Leon Vaganay and Ernest Cadman Colwell. The 
former roundly asserted that 'applied to New Testament texts 
this system is useless'. 1 The latter more cautiously declared that 
'the genealogical method is not of primary importance. . . . 
It can chart the history of transmission in an area narrowly 
limited in time and space. . . . But in the larger area where the 
larger questions are settled, it still has to demonstrate its value 
for the reconstruction of the original text of the Greek New 
Testament.' 2 

In evaluating the justice of Bedier's attack on the genealogical 
method, it ought to be pointed out that a much more innocent 
explanation lies behind the circumstance that almost all stem- 
mata result in two branches than the imputation of deliberate 
suppression or distortion of evidence. From the standpoint of 
mathematics, as Maas observes, 'We must remind ourselves that 
of the twenty-two types of stemma possible where three wit- 
nesses exist, only one has three branches.' 3 

Bedier's second criticism of Lachmannian methodology 
possesses a semblance of justification only when one treats 
manuscripts as though they were fixed entities, as unchanging 
as a printed book. On the contrary, after a manuscript was 
copied, it continued to live and was subject to modifications — 
as the numerous erasures, corrections, additions, glosses, and 
remarks by readers entered in the margins testify. One must 
therefore take into account what may be called the successive 
'stages of manuscripts', 4 as well as the possibility of multiple 

The disconcerting ambiguity arising from the construction 
of equally cogent classifications of the same manuscripts need 
not result in abandoning the Lachmannian method altogether. 

1 Leon Vaganay, Initiation a la critique textuelle neotestamentaire (Paris, 1 934) , p. 60 ; 
Eng. trans., p. 71. 

2 E. C. Colwell, 'Genealogical Method: its Achievements and its Limitations', 
Journal of Biblical Literature, lxvi (1947), p- 132. See also his article, 'The Significance 
of Grouping of New Testament Manuscripts', New Testament Studies, iv (1958), 

PP- 73-9 2 - 

3 Maas, op. cit., p. 48. 

* For a discussion of the implications of such alterations, marking successive 
stages in the history of the transmission of a treatise, see Jean Irigoin, 'Stemmas 
bifides et etats des manuscrits', Revue de philologie, lxxx (1954), pp. 211-17. 

Modern Methods of Textual Criticism 161 

Faced with a number of manuscripts which have a mixed 
ancestry, the textual critic may well adopt a compromise 
between the purely eclectic method advocated by Bedier and 
the rigid genealogical procedure of classical text criticism. Sup- 
pose that there are five manuscripts, A, B, C, D, and E, the 
lineage of which is not clear; we cannot say, for example, that 
ABC form one family, descended from a common ancestor, 
while D E form another. A comparison discloses, however, that 
certain characteristic readings are common to the group ABC, 
but are not present in D and E. This evidence shows that, so 
far as these readings are concerned, some manuscript with such 
characteristic readings was one member in the ancestry common 
to A, B, and C, though that ancestry may in other respects be 

One may conclude, therefore, that, despite the presence of 
a large amount of mixture in the ancestry of New Testament 
manuscripts, it will be advantageous for the textual critic to 
search out the broad features of more or less closely related 
groups of manuscripts. Such a process discloses that in general 
the Koine (or Byzantine) text of the New Testament is secondary, 
being characterized by the features which Hort delineated with 
classic vividness (see pp. 13 1-2 above). Moreover, major dis- 
locations common 'to the members of smaller groups of manu- 
scripts prove both the existence of such groups and their ultimate 
derivation from a common archetype which had suffered such 
dislocations (as, for example, the position of the pericope de adul- 
tera after Luke xxi. 38 in fam. 1 3) , I 


One of the axioms of classical textual criticism is brevior lectio 
potior, that is, the shorter of two readings is probably original. 
This principle, which has been accepted as generally valid by 
both classical and Biblical scholars, was challenged in 19 14 by 
Albert C. Clark in his inaugural lecture as Corpus Professor of 

1 Although the presence of mixture to a greater or less extent in New Testament 
manuscripts makes it impossible to draw up precise genealogical stemmata, the 
mixture itself carries with it compensatory benefits. As Zuntz has pointed out 
(Classica et Mediaevalia, iii [1940], p. 24), the more widely that cross-fertilization of 
manuscripts has taken place, the more improbable it is that any old reading, true or 
false, could have entirely disappeared. 

1 62 Modern Methods of Textual Criticism 

Latin in the University of Oxford. 1 Clark's researches in the 
manuscripts of Cicero's orations led him to believe that acci- 
dental omission was a much more common fault than deliber- 
ate interpolation by scribes. Four years later Clark published a 
lengthy treatise on The Descent of Manuscripts (Oxford, 1918), in 
which he showed that many omissions in classical texts involve 
multiples of the number of letters in an average line of script. Of 
two forms of text, one longer and one shorter, the latter can 
almost always be explained as the result of a scribe's omitting 
one or more lines of his exemplar. As Clark put it, 'A text is like 
a traveller who goes from one inn to another, losing an article of 
luggage at each halt.' 2 

Clark applied his principle, longior lectio potior, to the text of 
the Gospels and Acts, 3 with the result that the Western form of 
text, being in general the longer text, came off much better than 
it had at the hands of Westcott and Hort. If Hort could see 
no good in the Western text, Clark could see none in the Neutral 
text, which he regarded as the result of accidental omissions of 
multiples of lines of average length. 

Clark's theory of accidental scribal omissions was criticized 
on several scores by such eminent textual scholars as Sanday, 
Souter, and Kenyon. The latter pointed out, 4 for example, that 
(1) variation in the length of lines in manuscripts makes the 
method of counting letters unreliable except for very short 
passages; (2) accidental omissions would not account for the 
regular correspondence of the omissions with breaks in the sense ; 
(3) most of the variants involve, not omission, but differences in 
wording ; and (4) such narrow columns as Clark's theory neces- 
sitates are exceedingly rare in the early papyri (an argument 
that has become the stronger as more and more early papyri 
with relatively wide columns have come to light) . Furthermore, 
the circumstances of the transmission of the Gospel accounts were 
quite different from those of Cicero's Verrine orations. The 
Church preserved many traditions of the deeds and sayings of 

1 Albert C. Clark, Recent Developments in Textual Criticism, an Inaugural Lecture 
delivered before the University on 6 June 1914 (Oxford, 1914). 

2 Journal of Theological Studies, xvi (19,15), p. 233. 

3 Albert C. Clark, The Primitive Text of the Gospels and Acts (Oxford, 19 14). 

4 F. C. Kenyon, The Text of the Greek Bible (London, 1937), p. 231, and 'The 
Western Text in the Gospels and the Acts', Proceedings of the British Academy, xxiv 
(1938). PP- 287-315. 

Modern Methods of Textual Criticism 163 

Christ which had not been included in the Gospels (cf. John 
xxi. 25). It would be natural for these to slip into the text of the 
Gospels, either from the margins of other manuscripts or from 
the living memory of the Church. 

In a subsequent study of the Western text of the Acts, Clark 
returned to the controversy. 1 This time he practically aban- 
doned the theory of accidental omission and revived the sug- 
gestion proposed in the seventeenth century by Jean Leclerc, 
namely that Luke had himself produced two editions of Acts. 
This hypothesis shifts the inquiry from scribal transmission to 
deliberate editorial alteration on the part of the author or editor, 
and its validity must be tested on grounds other than those of 
textual criticism alone. The only comment which needs to be 
made here is that a comparison of the trends in the textual 
criticism of the Iliad and the Mahabharata, two great national 
epics the transmission of which reveals certain parallels to the 
transmission of the Gospels, is instructive for the New Testament 
scholar. Textual critics of both these corpora of quasi-religious 
literature are convinced that they are growing texts, and that no 
scribe deliberately excised any considerable portion of either 
poem. 2 


Several scholars have attempted to replace (or at least to 
supplement) the genealogical method by means of a purely sta- 
tistical analysis of variant readings. In his investigation of the 
transmission of the Latin Vulgate of the Old Testament, Dom 
Henri Quentin devised his so-called Rule of Iron {regie defer), 
according to which he submitted all variants to a rigidly 
mechanical method. Here is his own brief account of the process: 

At the very first I reject all thought of the primitive reading. I 
know neither errors, nor common faults, nor good readings nor bad, 
but only different forms of the text. Then, by a method resting upon 
the rigorous use of statistics, I discriminate family from family, 

« Albert C. Clark, The Acts of the Apostles (Oxford, 1933). 

2 For a survey of the trends of textual scholarship in the editing of the Iliad and 
the Mahabharata, reference may be made to B. M. Metzger, Chapters in the History of 
New Testament Textual Criticism (Leiden and Grand Rapids, 1963), pp. 142-54. 

164 Modern Methods of Textual Criticism 

classify the manuscripts composing each family, and finally classify 
the families themselves. From this classification results a critical 
canon that lays down an iron rule in the establishing of a text, and 
using this, I can reconstruct the archetype, which is the nearest form 
of text to the original that we can reach with extant manuscripts. 
Then, but only then, I allow myself to think of the original. I ex- 
amine the text from that point of view, and where the archetype is 
obviously faulty, I correct it by using the resources of internal criti- 
cism; always, however, taking care to indicate with a conventional 
sign that, at such and such a point, I have departed from the text 
resulting from the application of the critical canon. 1 

For the statistical examination of readings (at this stage he 
does not call them variants) Quentin devised an ingenious 
method which he called 'comparison by threes' (comparaison d 
trois). It aims at finding for every possible triad among the 
manuscripts of a text the one manuscript which is mediator 
between the other two. In order to determine whether in a 
given triad of manuscripts, as A, B, and C, any one is mediator 
between the other two, we merely count the agreements of 
A B against C, of A C against B, and of B C against A. If the 
figure for any of these relations is zero, the isolated manuscript 
is mediator between those against which it is opposed. Thus if 
we find that A C < B = o (which is to be read, 'A and C have 
no common deviation against B'), then B is mediator between 
A and C. This statistical examination, according to Quentin, 
gives clues for the grouping of manuscripts within a family, 
and, thanks to the overlapping of certain triads (those composed 
of manuscripts from different families), also for the relation- 
ship of larger groups. 

Through this method Quentin sought to reduce to three the 
basic manuscripts of the Vulgate from which the others are 
derived and which ought to make possible the reconstruction of 
the archetype. He was inclined to place this archetype about 
100 or 150 years after St. Jerome. For the Octateuch Quentin 

' Dora Henri Quentin, Essais de critique textuelle (Paris, 1926), p. 37. This volume 
as well as Quentin's earlier Mimoire sur V ' Hablissement du texte de la Vulgate (Rome and 
Paris, 192a) aroused a good deal of controversy; for a bibliography of appraisals of 
the method, see Paul Collomp, La Critique des textes (Paris, 1931), pp. 72 f., to which 
may be added J. Burke Severs, 'Quentin's Theory of Textual Criticism', English 
Institute Annual, 1941 (New York, 1942), pp. 65-93. One of the most trenchant 
criticisms is that of E. K. Rand in the Harvard Theological Review, xvii (1924), 
PP- 197-264. See below, p. 271. 

Modern Methods of Textual Criticism 165 

chose codex Turonensis of the Spanish group of manuscripts, 
codex Ottobonianus of the Theodulflan group, and codex 
Amiatinus of the pre-Alcuin group. Of these three he regarded 
the Tours manuscript, which contains the Pentateuch in a hand 
of the sixth or seventh century, as the best. 

The obvious criticism of the method as a whole is that it is too 
cumbersome to apply to a text which is of any considerable length 
and which is preserved in many manuscripts. In the former case 
obviously a selection must be made of sample passages. Quentin 
himself based his recension of the Octateuch on the evidence 
derived from no more than ninety-one groups of variants that 
were taken from eight sample chapters (one from each book). 
But such a procedure is open to the objection that the investi- 
gator has arbitrarily selected his evidence. Dom John Chapman, 
who examined some 2,000 variants from Genesis and Exodus, 
came to conclusions which are generally inconsistent with those 
of Quentin. 1 Furthermore, a purely mechanical examination of 
agreements of two manuscripts against a third must be checked 
by an examination of the nature of the relations which produce 
zeros — for the same omission by homoeoteleuton may occur in- 
dependently in the copying of two unrelated manuscripts. It is 
significant that Quentin's successors in the Benedictine project 
of editing the Vulgate Bible could not be persuaded to carry on 
his method beyond the first three books of the Old Testament, 
which Quentin himself had edited. 2 


Another proposal to edit texts by means of statistical analyses 
of variant readings was developed by Sir Walter W. Greg, a 
specialist in Middle English. While studying the relationships 
of manuscripts of the Chester miracle plays, Greg drew up a 
system which he called The Calculus of Variants: an Essay onTextual 
Criticism (Oxford, 1927). Like Quentin, Greg believed that it 
is possible to investigate the stemma of manuscripts without in- 
quiring into the question of the originality and non-originality 

1 Dom John Chapman, Revue Benidictine, xxvii (1925), pp. 6-46, 365-403. 

2 The Benedictine edition bears the title Biblia sacra iuxta latinam vulgatam 
editionem. For a description of the project, located at the newly founded Abbey of 
St. Jerome (Monastero San Girolamo) on the Janiculum Hill in Rome, see J. O. 
Smit, De Vulgaat (Roermond, 1948). 

826158 M 

1 66 Modern Methods of Textual Criticism 

of variant readings, operating simply on the basis of variational 
groups exhibited by the extant manuscripts. Using only logical 
deduction, Greg constructed an elaborate calculus of the possi- 
bilities of relationship among witnesses. Though his discussion 
is burdened by a needless proliferation of pseudo-mathematical 
symbols, Greg's volume is basically sound in theory. 

There is, however, a wide gap between Greg's theoretical 
consideration of the subject and the practical application of 
his principles to actual textual problems. In an attempt to lay 
down orderly and consistent rules for dealing with problems 
which Greg had deemed insoluble by logical methods, Archi- 
bald A. Hill developed what he called 'Some Postulates for 
Distributional Study of Texts'. 1 One of these postulates is the 
principle of simplicity, which Hill invokes in weighing the as- 
sumptions implied in alternate stemmata when there is more 
than one possible interpretation of the data. 

Adopting a distinction made by several scholars 2 between a 
distributional relationship among manuscripts (that having to 
do with a theoretical analysis of the statistics of combinations of 
manuscripts which share variant readings) and a bibliographi- 
cal relationship among manuscripts (that having to do with the 
actual chronological and genealogical filiation of manuscripts), 
Hill goes so far as to declare: 

A tree [i.e. stemma] is a description of the relationship of readings 
found in manuscripts, and ought never to be understood as a state- 
ment that A was copied from B. It merely states that the readings 
now found in A are derivable from readings now found in B, after 
examination of all the extant evidence. ... I am labouring the point, 
since students seem sometimes unnecessarily timid in the face of 
external evidence, particularly of chronology. . . . Actually an edi- 
tor who has derived A from B should be quite unmoved by external 
evidence that B is on twentieth century paper, while A is on medi- 
aeval vellum. 3 

To a scholar trained in the classical tradition, such a doctri- 
naire approach to the problems of textual criticism appears to 

1 Published in Studies in Bibliography: Papers of the Bibliographical Society of the 
University of Virginia, iii (1950-1), pp. 63-95. 

1 e.g. Edwin Wolf 2nd, ' "If Shadows be a Picture's Excellence" : an Experiment 
in Critical Bibliography', Publications of the Modern Language Association of America, 
lxiii (1948), pp. 831-57; cf. W. W. Greg, 'Bibliography — An Apologia', Library, 
4th ser., xiii (i93 2 )> PP- "3~43- 3 Op. cit., pp. 84 f. 

Modern Methods of Textual Criticism 167 

be tantamount to saying that when theories and facts disagree, 
so much the worse for the facts! 

The most recent development of the theoretical and practi- 
cal implications of Greg's calculus is found in the publications 
of Vinton A. Dearing, Associate Professor of English in the 
University of California at Los Angeles. Besides serving as tex- 
tual editor of the California edition of the works of John Dry- 
den ( 1 956-) , Dearing prepared a handbook of textual criticism 
setting forth a synthesis of Quentin's principle of mediators 
with Greg's calculus and Hill's principle of simplicity. In his pre- 
face Dearing candidly acknowledges: 

My method will not appeal at once to those who feel that 'common 
sense' is a sufficient guide to textual matters, or to those who are 
familiar with other methods. It is difficult to accept a new method 
for doing what is obvious or what an older method will do just as 
well. But any method is necessarily homogeneous. Because the older 
methods defy extension or development, a more comprehensive 
method necessarily presents alternate solutions to familiar problems. 
My method for the first time distinguishes the text conveyed by the 
manuscript — a mental phenomenon — from the manuscript convey- 
ing the text — a physical phenomenon. 1 

Confronted with these preliminary remarks, the reader is 
likely to pursue the subject with mixed feelings of curiosity and 

Besides setting forth a good deal of general textual methodo- 
logy, ranging from how to record variant readings to the pos- 
sibilities of using punched cards and computing machines for 
statistical analysis and comparison of manuscripts in textual 
criticism, Dearing deals with several text-critical problems, in- 
cluding an analysis of the Greek text of Paul's Epistle to Phile- 
mon — the end for which, so the author states, the volume was 
initially designed. As regards the textual evidence of Philemon, 
Dearing confines his attention to the ten uncial manuscripts 
cited by Tischendorf in his eighth edition. 2 These ten manu- 
scripts present fourteen different 'states' of text, inasmuch as 
four of them contain later scribal alterations. After tabulating 

1 Vinton A. Dearing, A Manual of Textual Analysis (Berkeley and Los Angeles, 
!959)> PP- viiif. 

2 Dearing takes no account of the variant readings from forty-two minuscule 
manuscripts cited by Tischendorf for Philemon. 

1 68 Modern Methods of Textual Criticism 

all the different combinations of states of manuscripts which 
support the variant readings, the author constructs what he calls 
a textual scheme, concluding that the text of the ancestor of 
the fourteen states of the manuscripts can be established 'when 
D* or E* or both agree with K c or P and when K c and P 
agree together'. 1 On the basis of such a purely mathematical 
methodology in dealing with the evidence, Dearing declares 
that in his opinion Tischendorf's text agrees with the hypo- 
thetical ancestor except that the latter read Xpiorov 'Iijuow in 
verse 6, Sea/Mis imv in verse 10, 6V aoi (without /cat) in verse 1 1, 
ai> 8e avrov and anXdyxva, irpooXafiov in verse 12, and ap.rjv in 
verse 25. 

The New Testament textual critic will doubtless find Dear- 
ing's application of the statistical method to the Greek text of 
Philemon an interesting experiment, but he may well wonder 
whether the results are worth the effort to achieve them, parti- 
cularly in view of Dearing's repeated declaration that he has 
been concerned to construct a textual, not a bibliographical, 
scheme. 2 That is, Dearing moves in the realm of 'states' of 
manuscripts, and his findings may or may not coincide with the 
actual chronological descent of the manuscripts. 

By way of concluding his discussion of the variants selected 
from Tischendorf's apparatus, Dearing writes: 'In time the full 
collations necessary for a more final analysis will be available, 
and the necessary mechanical aids to calculation will have been 

1 Dearing, op. cit., p. 93. 

1 Dearing's distinction between bibliographical and textual stemmata (compare 
Hill's distinction between distributional and bibliographical trees, p. 166 above) 
can be illustrated by the following example: 'Suppose that manuscript B has been 
copied from manuscript A without verbal alteration, and that manuscript C has 
been copied from A with some verbal variants. Now, there are two ways of inter- 
preting the relationship of C to B. From a purely mechanical point of view, C does 
not in any sense derive from B, since both derive independently from A. On the 
other hand, the readings of C could be said to derive from those of B. Moreover, if 
only B and C are extant, the textual critic may be unable to distinguish B from A, 
and he may find it convenient, as well as harmless, to assume that B is A and there- 
fore an ancestor of C. The first point of view Mr. Dearing labels "bibliographical", 
the second "textual". In somewhat the way mathematical laws are used in physics, 
this distinction permits the critic to work out a stemmatic diagram on the abstract 
"textual" level before returning to the factual "bibliographical" level for final estab- 
lishment of text. Such procedure should be valid if there is no attempt to draw un- 
justified "bibliographical" inferences from "textual" assumptions' (from David M. 
Vieth's review of Dearing's Manual in Journal of English and Germanic Philology, lix 
[i960], p. 556). Cf. Dearing's subsequent study, 'Some Notes on Genealogical 
Methods in Textual Criticism', Novum Testamentum, ix (1967), pp. 278-97. 

Modern Methods of Textual Criticism 169 

provided. That happy day, it is to be hoped, is not far in the 
future.' 1 The present writer confesses that he does not share such 
sanguine expectations of benefits to be derived from what Dear- 
ing regards as 'necessary mechanical aids'. A mechanical pro- 
cedure which dispenses with the subjective element in criticism 
reminds one of George Foot Moore's trenchant remark that 'the 
methodical elimination of the element of human intelligence 
can hardly be the ideal of science'. 2 Though computing machines 
may conceivably be useful in 'remembering' the statistical de- 
tails of variant readings, it is not likely that they will replace the 
use of rational critical processes in evaluating 'good' and 'bad' 
readings. 3 


During the past generation a British churchman and an 
Italian classicist made significant contributions to textual criti- 
cism. Though their areas of interest were widely different, in 
some respects the methods which the two scholars proposed are 


In 1924 Canon Streeter published a volume on The Four 
Gospels, a Study of Origins, in which solid scholarship is com- 
bined with a fertile imagination and an engaging literary style. 

1 Op. cit., p. 93. For a description of the use of mechanical aids in textual editing, 
see the pamphlet entitled Methods of Textual Editing, a paper delivered by Vinton A. 
Dearing at a Seminar on Bibliography held at the Clark Library, 12 May 1962 
(University of California, Los Angeles), pp. 18 ff. Cf. also John William Ellison, 
'The Use of Electric Computers in the Study of the Greek New Testament Text' 
(Unpublished Diss., Harvard University, 1957). See below, p. 271. 

2 Moore's preface to Henry St. John Thackeray's Josephus, the Man and the 
Historian (New York, 1929), p. v. 

3 One of the curiosities of statistical pseudo-scholarship (which need not be 
dignified beyond this notice in a footnote) is Ivan Panin's The New Testament in the 
Original Greek, the Text Established by Means of Bible Numerics, privately printed by 
John Johnson at the University Press, Oxford, 1934. Panin prepared his edition 
by making such alterations in Westcott and Hort's text as he thought were justified 
on the basis of 'Bible Numerics', a system which he says he discovered in 1890. 
According to this 'method' the editor gives preference to that variant reading which 
yields the greatest number of multiples of 7, 11, 15,23, &c, in the number of vowels, 
consonants, syllables, parts of speech, &c, in a given verse or pericope. For a 
refutation of the system (if refutation be deemed necessary!), see Oswald T. Allis, 
Bible Numerics (Chicago, 1944), and, more briefly, J. Oliver Buswell, 'Bible 
Numerics — the True and the False', Moody Monthly, xxxii (1932), pp. 530-1, and 
'Notes on Open Letters', Sunday School Times, lxxxiv (1942), pp. 1058, 1072-3. 

170 Modern Methods of Textual Criticism 

Building on Westcott and Hort's classic work, Streeter refined 
their methodology in the light of the acquisition of new manu- 
script evidence since 1881. Adopting an idea which Hug had 
first developed, Streeter emphasized the importance of isolating 
the forms of text which were current at the great centres of 
ancient Christianity. By means of evidence derived from quota- 
tions in the writings of early Church Fathers, he isolated and 
identified the characteristic forms of New Testament text which 
had developed at the principal sees of the ancient Church. By 
about a.d. 200 these local texts had reached (so Streeter believed) 
their maximum divergence, a divergence which is reflected in 
the earliest Syriac, Latin, and Coptic versions. It is probable 
that the oldest forms of these three versions were derived re- 
spectively from the Greek texts current in Antioch, Rome, and 

Besides these three forms of text, Streeter's analysis of the 
evidence of codex Koridethi (0) and some of the writings of 
Origen and Eusebius led him to postulate the existence of a 
so-called Caesarean text of the Gospels, to which fam. 1 and 
fam. 1 3 also belong. Streeter combined into one text-type, which 
he designated the Alexandrian, the witnesses that Westcott and 
Hort had assigned to their Neutral and Alexandrian groups. He 
agreed with Westcott and Hort that the Syrian text, which he 
renamed the Byzantine text, arose during the fourth century 
through the recensional activity of Lucian of Antioch and was 
adopted by about 380 at Constantinople. This text became the 
prevailing ecclesiastical form of the New Testament throughout 
the Greek-speaking world, and eventually constituted the basis 
of the Textus Receptus. Therefore readings later than the fifth 
century, Streeter argued, can be ignored except when they dif- 
fer from the prevailing Byzantine text. On the other hand, be- 
cause it is possible that an ancient form of text may have been 
preserved at a relatively late date in a locality cut off from the 
main stream of Christianity, the precedence of manuscripts 
depends not so much on their age as on their pedigree. 

The relationship between the local texts used in the five 
Churches of Alexandria, Caesarea, Antioch, Italy and Gaul, and 
Carthage stands in a graded series corresponding to their geo- 
graphical propinquity around the eastern Mediterranean world. 
Streeter sets forth this point as follows: 

Modern Methods of Textual Criticism 171 

Each member of the series has many readings peculiar to itself, but 
each is related to its next-door neighbour far more closely than to 
remoter members of the series. Thus B (Alexandria) has much in 
common with Jam. & (Caesarea) ;fam. & shares many striking read- 
ings with Syr. S. (Antioch) ; Syr. S. in turn has contacts with D b a 
(Italy-Gaul) ; and, following round the circle to the point from 
which we started, k (Carthage) is in a sense a half-way house be- 
tween D b a and B (Alexandria again). 1 

Figs. 3 and 4 reproduce a stemma and a chart from Streeter's 
volume, showing the relationship of the several local texts and 
the chief witnesses which support each. 

Original Autographs 





Revised Text of Lucian c.A.D.310 
Byzantine (or 'Standard' ) Text. (A) E etc. etc. 


W Mkfc e 

Textus Receptus 

Fig. 3. Stemma Illustrating Streeter's Theory of Local Texts 
(from B. H. Streeter, The Four Gospels, p. 26) 

Some of the practical conclusions which follow from the ac- 
ceptance of Streeter's theory of local texts include the follow- 
ing, which are set forth in his own words. 

(a) The textual critic, in weighing the amount of external evidence 
in favour of any reading, should consider primarily, not the number 
or the age of the manuscripts which support it, but the number and 
geographical distribution of the ancient local texts in which it can 
be traced. 

1 B. H. Streeter, The Four Gospels, p. 106. 


Modern Methods of Textual Criticism 























d ^.^ 



2 e 



O 00 



- » 



- 8 


(3 3 

•2 S 





"S a 






S 1 * 

~ 9 

5: « 


« O 

o ^ 
i-i ^ 



9° 3 














(2 a 







03 03 




o o 



i j 


C5 C3 







3 S ' N 




















c 1 


















s i 

CO ^. 

I ^". 

tO r . OP 

co ti w 



— >i J3 

I o JS> 



^ -- o 

* c: TS 
n ^ *2 



a M in 

J. st 


! o <2, 

J. r*! 
2 oj . 


• i v 

Modern Methods of Textual Criticism 173 

(b) It follows that manuscripts should be cited, not in alphabetical 
or numerical order, but in groups corresponding to the local texts 
which they represent. When at least three of the leading representa- 
tives of any local text support a reading, very little is gained by 
citing additional evidence of manuscripts which normally support 
the same local text. 1 

(c) Though on minor points of reading absolute certainty may 
often be unattainable, a text of the Gospels can be reached, the 
freedom of which from serious modification or interpolation is 
guaranteed by the concurrence of different lines of ancient and 
independent evidence. 2 


In 1932 the learned Italian classicist, Giorgio Pasquali, pub- 
lished a monumental discussion of problems involved in the 
editing of ancient Greek and Latin authors. 3 The author, who 
had previously devoted more than forty pages to a review of 
Paul Maas's Textkritik, a succinctly written treatise of eighteen 
pages, 4 casts his net wide and illustrates his arguments by 
thousands of references to ancient authors and modern editors. 
Despite Pasquali's almost exclusive preoccupation with non- 
Biblical texts, the New Testament scholar will find that more than 
one suggestion in the volume bears indirectly upon his own 
special problems. Some of the points which Pasquali makes are 
the following: 5 

(a) The medieval manuscripts of Greek and Latin authors 
often go back, directly or indirectly, to ancient editions that 
already differed among themselves. 

(b) Coincidence in obvious errors or in 'trivializzazioni' does 
not prove relationship between manuscripts. Furthermore, in 
general the coincidence of readings in diverse witnesses need not 
prove relationship, because original readings can be preserved 
independently in several branches of the tradition. 

(c) For Latin authors, whose manuscript tradition is much 
richer than that of most Greek authors, a codex recentior is not 

1 B. H. Streeter, op. cit., p. 78/ 2 Ibid., p. 148. 

3 Giorgio Pasquali, Storia delta tradizione e critica del testo (Florence, 1934 ; 2nd ed., 


4 The review appeared in Gnomon, v (1929), pp. 417-35, 498-521. 

5 Pasquali sets them forth conveniently in the preface of his volume, pp. xv ff. 

1 74 Modern Methods of Textual Criticism 

necessarily a codex deterior. The authority of a witness is inde- 
pendent of its antiquity. 

(d) The collations of classical authors made by humanist 
scholars of the Renaissance, as well as the editions printed at 
that time, rest in part upon manuscripts which are now lost. 
Therefore a unique variant in a Renaissance text is not always 
to be credited to the ingenuity of contemporary editors, but may 
go back to very ancient times. 

(e) Arbitrary modifications introduced by a scribe into a 
manuscript may cause us to doubt the value of the whole, but they 
do not entitle us a priori to neglect its authority at other points, 
on the pretext that it is merely an interpolated manuscript. 

(/) There is a widespread prejudice which believes that 
the transmission of manuscripts has been purely mechanical, 
the work of 'stupid' scribes who transcribed without thinking. 
There were other scribes, however, who embellished, simplified, 
and adapted or even collated texts with other documents. A 
conjecture which is palaeographically justified may enable us 
to recover, in the case of mechanical transmission, the original 
reading. On the other hand, in an 'open' recension where 
'intelligent' scribes have intervened, internal criticism plays a 
much more important role in the proposing of emendations. 

(g) There is a widespread prejudice which believes that the 
transmission of texts is solely 'vertical' and in chronological 
sequence; in some cases, however, it has taken place also on a 
horizontal or transverse plane. The collation of a manuscript 
with other manuscripts besides the one which served as the 
exemplar has introduced readings from one branch into another, 
thus diversifying and contaminating the text. 

(h) In general linguistics it is commonly acknowledged that the 
most ancient stages of a phenomenon are preserved best in the 
peripheral zones of the area of their dispersion, and the co- 
incidence of two forms current in regions far removed from the 
centre and from each other constitutes a presumption in favour 
of their antiquity. In the same way the coincidence of particular 
readings in manuscripts of divergent origin often proves their 

In these paragraphs, it will have been observed, Pasquali 
enunciates principles new and old. Without referring to Streeter's 
theory of 'local texts' of the New Testament, he suggests that 

Modern Methods of Textual Criticism 175 

something similar may have occurred in the case of several an- 
cient classical authors. Furthermore, he is inclined to think that, 
at least as regards the Homeric tradition, Byzantine manuscripts 
(which may go back to ancient editions) can be relied upon 
to preserve a text of more value than that contained in papyri, 
which, though earlier, represent an uncontrolled 'wild' text. 

Perhaps one of the more significant points in Pasquali's 
volume which may aid the New Testament critic is his reference 
to what is popularly called 'linguistic geography' (the final item 
mentioned above) . The research of Bartoli and other scholars 1 
on the diffusion of dialectal forms of speech has not a little to teach 
the textual critic concerning the somewhat related problems of 
the dissemination of variant readings. 


Dissatisfied with the results achieved by weighing the external 
evidence for variant readings in terms of support from individual 
manuscripts or families of manuscripts and local texts, several 
scholars have directed primary attention to the individual 
variants themselves in an effort to find which will account best 
for the rise of the others. This process has been given various 
names. It has been called eclecticism, 2 because in its application 
the textual critic pays less attention to questions of date and 
families of manuscripts than to internal or contextual considera- 
tions. Consequently the editor of a text follows now one and now 
another set of witnesses in accord with what is deemed to be the 
author's style or the exigencies of transcriptional hazards. The 
ultimate disregard for the external evidence of variants is to be 
seen in several of Paul Mazon's editions of the Greek clas- 
sics, where the apparatus supplies only the variants without 
mentioning the identity of the witnesses which support them. 3 

1 Matteo Bartoli, Introduzione alia neolinguistica (Geneva, 1925), and Saggi di 
linguistica spaziale (Turin, 1 945) . Several atlases of dialects have been or are in the 
process of being compiled ; for a list, see Jos. Schrijnen, Essai de bibliographie de glo- 
graphie linguistique gintrale (Nimegue, 1933), and for a recently published atlas, see 
K. Heeroma's Taalatlas van Oost-Nederland en aangrenzende gebieden (Assen, 1 957-) . 

2 e.g. by G. D. Kilpatrick, Journal of Theological Studies, xliv (1943), p. 36 
('rigorous eclecticism'), and ibid, xlv (1944), p. 65 ('impartial eclecticism'). See 
below, p. 271. 

3 Cf. A. Dain's comments on 'Olympian indifference' to what is regarded as 
'useless erudition' (Les Manuscrits [Paris, 1949], p. 161 ; rev. ed. [1964], p. 175). 

1 76 Modern Methods of Textual Criticism 

Another descriptive name which has been given to this proce- 
dure of handling the textual evidence is rational criticism. 1 The 
use of the adjective 'rational' in this connexion is not intended 
to suggest that all other methods of criticism are irrational, 
but that the critic is concerned primarily with finding plausible 
reasons based on internal considerations to justify in each case 
his choice of one reading as original and the others as secondary. 

It will be understood that almost all textual critics have paid 
attention to aspects of rational criticism. Certain scholars, 
however, have given primary and sometimes exclusive con- 
sideration to the style of the author and the demands of the 
context. Such, for example, was the method which Bernhard 
Weiss followed in preparing his edition of the Greek Testament 
(see pp. 137 f. above). Likewise C. H. Turner, having made a 
thorough study of Marcan usage, reconstructed the Greek text 
of the first chapter of Mark in accord with stylistic considera- 
tions. 2 One of the conclusions to which Turner's investigations 
led him was the need to show greater respect for Western 
readings, though they may be supported by only a few witnesses 
(for instance, by D and one of the three leading Old Latin manu- 
scripts, k or e or a). 3 

More recently several scholars have examined the text of 
certain other New Testament books in the light of rational 
criticism. For the Pauline Epistles, and especially for 1 Corin- 
thians and Hebrews, the lively and learned Schweich Lectures 
delivered in 1946 by Giinther Zuntz have much to teach the 
student regarding text-critical method. 4 As regards the Book of 

1 M.-J. Lagrange uses this term in the title of his monumental volume, Critique 
textuelle; ii, La Critique rationnelle (Paris, 1935), though in practice he frequently 
pays more attention to external evidence than one would have thought likely. 

2 C. H. Turner, 'A Textual Commentary on Mark 1', Journal of Theological 
Studies, xxviii (1926-7), pp. 145-58, and 'Marcan Usage : Notes, Critical and Exege- 
tical, on the Second Gospel', ibid, xxv (1923-4), pp. 377-86; xxvi (1924-5), pp. 12- 
20, 145-56, 225-4°. 337-46; xxvii (1925-6), pp. 58-62; xxviii (1926-7), pp. 9-30, 

3 See his article, 'Western Readings in the Second Half of St. Mark's Gospel', 
Journal of Theological Studies, xxix (1927-8), pp. 1-16, and 'The Textual Criticism 
of the New Testament', in A New Commentary on Holy Scripture . . ., ed. by Charles 
Gore el al. (London, 1928), pp. 718-29. For adumbrations of this point of view, 
compare Turner's earlier article, 'New Testament, Text of, in The Illustrated Bible 
Dictionary, ed. by William C. Piercy (New York, 1908), pp. 585-96. 

4 G. Zuntz, The Text of the Epistles: a Disquisition upon the Corpus Paulinum 
(London, 1953). 

Modern Methods of Textual Criticism 177 

Revelation, one of the most valuable sections of Josef Schmid's 
extensive monograph on the history of the Greek text of the 
Apocalypse is his chapter dealing with the bearing of the linguistic 
usage of the book on the textual evaluation of variant readings. 1 
Turner's reliance on stylistic criteria in making textual de- 
cisions has been revived by George D. Kilpatrick of Queen's 
College, Oxford. Portions of his research into questions of lexical 
and grammatical usage of authors of the New Testament have 
been published in several periodicals, 2 and his text-critical con- 
clusions are finding expression in the fascicles of A Greek-English 
Diglotfor the Use of Translators, issued for private circulation by 
the British and Foreign Bible Society {Mark, 1958; Matthew, 
1959; John, i960; The General Letters, 1961 ; Luke, 1962). 3 Of two 
or more variant readings, Kilpatrick usually prefers that one 
which accords with what is deemed to be the author's style, 
irrespective of the date and nature of the external evidence 
supporting the reading. In matters on which no firm decision 
can be made concerning the author's style, he often appeals to 
the criterion of Atticism, which became one of the dominant ten- 
dencies in literary circles during the first and second Christian 
centuries. He argues that scribes in the second century intro- 
duced many Atticisms into the text of the New Testament. 4 Of 
two readings, therefore, one of which conforms to Attic canons 
and the other does not, he is inclined to accept the non-Attic 
reading, even though no early manuscript evidence may support 
it. In order to justify his general disregard for the age and quality 
of external evidence, Kilpatrick declares that by about a.d. 200 
the great majority of the deliberate changes had been introduced 
into the textual stream, and that thereafter scribes transmitted 
the several forms of text with great fidelity. Thus, though a 
variant reading may happen to be preserved only in a late 
minuscule manuscript, if it is in harmony with what is taken to 

1 Josef Schmid, Studienzur Geschichte des griechischen Apokalypse-Textes; ii, Die alten 
Stdmme (Munich, 1955), pp. 173-251. 

2 G. D. Kilpatrick, 'Some Notes on Marcan Usage', Bible Translator, vii (1956), 
pp. 2-9, 51-56, 146; 'Some Notes on Johannine Usage', ibid, xi (ig6o), pp. 173-7 '> 
and 'SiaXeyeadai and hia\oyil,<LaBa.i in the New Testament', Journal of Theological 
Studies, n.s., xi (i960), pp. 338-40. 3 See below, p. 271. 

4 See Kilpatrick's article, 'Atticism and the Text of the Greek New Testament', 
in Neutestamentliche Aufsdtze: Festschrift fur Prof. Josef Schmid (Regensburg, 1963), 
pp. 125-37; compare also his comments in Goltingische gelehrte Anzeigen, ccxv 
(■9 6 3). PP- Hff- See below, p. 271. 

178 Modern Methods of Textual Criticism 

be the author's style or reflects a non-Atticistic tendency, Kil- 
patrick is disposed to regard it as original. 

The extent to which Kilpatrick is prepared to go in adopting 
readings which have the most meagre external support, if he is 
convinced that internal considerations require it, may be illus- 
trated by the following readings in the Greek-English Diglot: 

Matt. xx. 30 eKpa&v is supported by 1 18 209 Syr c - P al . 

Matt. xxii. 1 omit eiirev with E SyrP. 

Matt. xxii. 7 aKovcras Se o flacriAevs ixeivos is supported by 33. 

Mark v. 11 re oprj is supported by 372 485 Syr s . 

Mark ix. 1 7 cmoKpidds avr& is supported by G. 

Mark xiv. 6 els ipe is supported by 517 579 /251 Syr s - p Eth. 

Mark xiv. 31 iXdXei. n&XXov is supported by 574 and k. 

Luke ix. 51 icrrqpi^ev is supported by 1241. 

John xix. 35 aX-qd-qs is supported by X 124 Ghr. 

Jas. ii. 18 epyaiv tt]v rricrnv jxov is supported by Syr h . 

1 Pet. ii. 1 1 cmixeadai v/ is supported by Vulg Gyp. 

2 John, vs. 8 7rAi)pi)? is supported by L Dam. 

By way of summary, it is obvious that there is much to com- 
mend the practice of a judicious eclecticism in text criticism, 
for no one manuscript and no one family preserves the original 
text in its entirety. And certainly the critic must always take into 
account what is or is not the usage of a given author, for, as 
A. E. Housman once expressed it in a pithy epigram: 'The 
indulgence of love for one manuscript and dislike for another 
inevitably begets indifference to the author himself.' 1 At the 
same time, however, the weaknesses inherent in the method 
should not be overlooked. Statistics regarding an author's usage 
are sometimes derived from concordances which are based on 
editions of the Greek Testament that contain, in some passages, 
quite indefensible readings. Furthermore, even if evidence re- 
garding an author's usage has been sifted critically, its signifi- 
cance for the passage at hand must be weighed in the light of 
two possibilities: (a) an author may on occasion vary his usage, 
and (b) a scribe who was aware of the author's prevailing usage 
may have altered a reading in order to bring it into harmony 
with that usage. 

One may conclude, therefore, that although considerations of 
the literary usage of a New Testament author will be of con- 
1 In Housman's edition of Lucan's Bellum civile (Oxford, 1926), p. vi. 

Modern Methods of Textual Criticism 1 79 

siderable value to the textual critic, it must not be made the 
primary criterion in the evaluation of variant readings to the 
virtual neglect of external evidence. Furthermore, to apply 
rigorously considerations based on the Atticistic revival during 
the early Christian centuries is to be in danger of disregarding 
the operation of other literary and stylistic tendencies that also 
influenced the Koine during the same period, some of which 
were deliberately anti- Atticistic. 1 


Obviously the most satisfying method of locating a newly 
discovered manuscript within the manuscript tradition of the 
New Testament is to compare it with each and all previously 
known manuscripts. This procedure, however, entails such an 
enormous amount of labour, out of all proportion to the impor- 
tance of the results, that scholars are content to make analyses 
on a selective basis. Sometimes the manuscript is compared 
with individual manuscripts which are typical of several fami- 
lies or types of text. The disadvantage of this procedure is that 
it ignores a large part of the evidence, and consequently one's 
conclusions may be upset when comparison is made with addi- 
tional manuscripts. 

More often the manuscript is collated against the Textus 
Receptus and then the variants are analysed in terms of agree- 
ment with a large number of manuscripts whose readings have 
been reported in various apparatus critici. Such a method of com- 
parison, however, is scarcely more satisfactory than the other, 
for it, too, omits a significant portion of the evidence, namely those 
readings which the manuscript shares with the Textus Receptus. 
Furthermore, analyses of a text in terms of differences from the 
Textus Receptus are of little value when no control has been 
provided. Thus it is quite deceptive to know only that in a 

1 See, for example, the references under 'Antiatticismus' in Wilhelm Schmid's 
Der Atticismus in seinen Hauptuertretern (Stuttgart, 1887-96), and compare Rader- 
macher's warning against undue reliance upon statistical considerations in judging 
the literary style of an author in a period when quite opposite influences cut across 
one another ; Ludwig Radermacher, Koine ( = Sitzungsberichte der Akademie der Wis- 
senschaften in Wien, Phil.-hist. Kl., ccxxiv, 5. Abhandlung, 1947), pp. 61 f. 

180 Modern Methods of Textual Criticism 

certain chapter a given manuscript agrees with, say, B and X 
ten times in differing from the Textus Receptus ; for if B and X 
should also, in that chapter, differ from the Textus Receptus in 
ninety other instances, the Alexandrian element in the given 
manuscript would be slight indeed. 

In order to provide a more exact method of analysis of manu- 
script relationships, as long ago as 191 1 E. A. Hutton proposed 
the use of what he called 'Triple Readings' in the analysis of 
textual affinities. 1 Selecting a list of readings in which the 
Alexandrian, Western, and Syrian (Byzantine) authorities are 
divided, he urged that henceforth all manuscripts be analysed 
by determining the number of agreements with each of these 
three textual types. 

With the multiplication of the number of identifiable textual 
groups, it is desirable to seek a higher degree of precision than 
Hutton's method permits. E. C. Colwell, with the assistance of 
M. M. Parvis, elaborated what is called the method of Multiple 
Readings in determining the relationship of manuscripts. A 
Multiple Reading is defined as 

one in which the minimum support for each of at least three variant 
forms of the text is either one of the major strands of the tradition, or 
the support of a previously established group (such as Family 1, 
Family 77, the Ferrar Group, K 1 , K', K r ), or the support of some 
one of the ancient versions (such as af, it, sy s , sy c , bo or sa), or the 
support of some single manuscript of an admittedly distinctive 
character (such as D). 2 

Colwell applied these specifications for Multiple Readings to 
a sample section of text extending from John i. 1 to iv. 40, and 
found twenty-two instances of Multiple Readings. By tabulating 
the number of agreements of a newly found manuscript with 
the several witnesses at those passages which involve Multiple 
Readings, one will learn something of the textual complexion 
of that manuscript. 

1 Edward Ardron Hutton, An Atlas of Textual Criticism, being an Attempt to Show 
the Mutual Relationship of the Authorities for the Text of the New Testament up to about 
1000 A.D. (Cambridge, 191 1). 

1 E. G. Colwell, 'Method in Locating a Newly-discovered Manuscript within 
the Manuscript Tradition of the Greek New Testament', Studia Evangelica . . ., 
ed. by Kurt Aland et al. ( = Texte und Untersuchungen, lxxiii, Berlin, 1 959) , p. 759 ; cf. 
also G. D. Dicks, Journal of Biblical Literature, lxvii (1948), pp. 366-8. 

Modern Methods of Textual Criticism 181 

The limitation of this method, at least as applied to the sample 
from John, is that test passages are relatively few in number 
in comparison with the extent of text investigated. It is con- 
ceivable that a given manuscript may agree consistently with 
a particular type of text in most of the twenty-two passages 
in the 152 verses covered by John i. 1 to iv. 40, but may 
disagree with that text-type in a significant number of read- 
ings in the remaining 130 verses not represented by Multiple 

It will be apparent from the criticisms levelled at each of the 
methods mentioned above that no single method thus far devised 
is entirely satisfactory in determining the textual complexion of 
a New Testament manuscript. The area tested must be large 
enough to be representative of the whole. At the same time the 
method must be sensitive enough to disclose whether various 
sections of a manuscript have been transcribed from exemplars 
of diverse text-types. It has been suggested elsewhere by the 
present writer that in analysing the textual complexion of 
lectionaries, which are normally collated against the Textus 
Receptus, it is necessary to have a control to determine precisely 
what percentage of Alexandrian, Western, and Caesarean 
deviations from the Textus Receptus are present in sample 
lections. 1 Furthermore, in all analyses the textual critic must 
give attention to the problem which arises when one attempts to 
assess the quality as well as the quantity of variants. The possi- 
bility that chance and not genealogical descent may account for 
the agreement of manuscripts in certain small variations — such 
as the presence or absence of the article with a proper name, or 
the aorist for the imperfect tense, and vice versa — has not been 
sufficiently taken into account. 2 

1 This method entails first the construction in a sample passage of a 'pure* 
Alexandrian text, a 'pure' Western text, a 'pure' Caesarean text, a 'pure' text of 
fam. 13, &c., and then the collation of each of these 'pure' texts against the Textus 
Receptus. By comparing the total number of variants from the Textus Receptus 
in each of these 'pure' texts with the number of variants which are present also in the 
Lectionary text, one can determine with precision the degrees of relationship. For an 
example of the application of this method of textual analysis, reference may be made 
to the present writer's monograph, The Saturday and Sunday Lessons from Luke in 
the Greek Gospel Lectionary (Chicago, 1944), pp. 24 ff. 

2 See, for example, the discussion of 'Quantita e qualita delle varianti', in Paolo 
Sacchi, Alle origini del Nuovo Testamento: Saggio per la storia della tradizione e la critica 
del testo (Firenze, 1956), pp. 86 ff. See below, p. 272. 

826158 N 

1 82 Modern Methods of Textual Criticism 


The classical method of textual criticism regularly involves, 
as was mentioned at the beginning of this chapter, the exercise 
of conjectural emendation. If the only reading, or each of several 
variant readings, which the documents of a text supply is im- 
possible or incomprehensible, the editor's only remaining re- 
source is to conjecture what the original reading must have been. 

A typical emendation involves the removal of an anomaly. 
It must not be overlooked, however, that though some anoma- 
lies are the result of corruption in the transmission of the text, 
other anomalies may have been either intended or tolerated by 
the author himself. 1 Before resorting to conjectural emenda- 
tion, therefore, the critic must be so thoroughly acquainted with 
the style and thought of his author that he cannot but judge 
a certain anomaly to be foreign to the author's intention. 

This aspect of criticism has at times been carried to absurd 
extremes. In his later work Richard Bentley, for example, 
largely disregarded the evidence of manuscripts in determining 
the correct readings, and depended chiefly upon his own in- 
stinctive feeling as to what an author must have written. He 
justified such a procedure in the magisterial phrase, nobis et 
ratio et res ipsa centum codicibus potiores sunt, which may be ren- 
dered 'for me both reason and the subject-matter are worth more 
than a hundred manuscripts'. In following this bold principle 
he did much that was rash and indefensible as well as much that 
is brilliant and convincing. The reductio ad absurdum of such a 
subjective method is found in Bentley's edition of Milton's 
Paradise Lost, in which he offers more than 800 emendations, 
restoring what in his opinion Milton must have really said (or 
meant to say) while dictating the poem to his daughters. 2 

Before a conjecture can be regarded as even probable, it must 
satisfy the two primary tests which are customarily applied in 
evaluating variant readings in manuscripts: (1) it must be 
intrinsically suitable, and (2) it must be such as to account for 
the corrupt reading or readings in the transmitted text. There 

1 For a discussion of the paradoxical possibility of a textual critic's 'improving' 
on the original, see G. Zuntz's article on I Cor. vi. 5 entitled 'The Critic Correcting 
the Author', Philologus, xcix (1955), pp. 295-303. 

1 See James Henry Monk, The Life of Richard Bentley, D.D., 2nded.,ii (London, 
1833), pp. 309-23, and Richard C. Jebb, Bentley (London, 1889), pp. 180-91. 

Modern Methods of Textual Criticism 183 

is, however, an important difference between the method of 
applying these tests to a conjectural emendation, and that of 
applying them to variants in manuscripts. We accept the variant 
which best satisfies the tests ; but we require of a successful con- 
jecture that it shall satisfy them absolutely well. The conjecture 
does not rise from a certain level of probability ('a happy guess') 
to the level of certainty, or approximate certainty, unless its 
fitness is exact and perfect. The only criterion of a successful 
conjecture is that it shall approve itself as inevitable. Lacking 
inevitability, it remains doubtful. 

An example from English literature will illustrate the wide dif- 
ferences of merit among proposed conjectures. 1 Since the early 
printers in England were often foreigners, who made quite as 
bad mistakes as their predecessors the scribes, the text of Shake- 
speare contains almost as many problem passages as that of 
Aeschylus. In the folio editions of Henry V, Act 11, scene iii, the 
hostess says of the dying Falstaff, 'his nose was as sharp as a 
pen and a table of Green Fields'. The words 'a table of Green 
Fields', which appear with trifling variations of spelling in the 
folio editions but which are omitted in the quarto editions, have 
been the subject of numerous conjectural emendations. Pope 
suggested (perhaps ironically) that this was a stage direction to 
bring in one of Greenfield's tables, Greenfield being supposed 
to be the furniture-dealer who supplied props for Shakespeare's 
theatre. Collier proposed 'on a table of green frieze', and another 
critic suggested 'or as stubble on shorn fields'. The conjecture 
which today is adopted by editors is 'and a' babbled of green 
fields', being a modification by Theobald of a happy proposal 
made by an anonymous annotator who corrected 'a table' to 
'a' talked'. 2 

The fault most often committed in the use of conjectural 

1 This example is taken nearly verbatim from James Gow's Companion to School 
Classics, 2nd ed. (London, 1889), pp. 65 f. See below, p. 272. 

2 Several passages in Shakespeare are corrupt beyond the ingenuity of palaeo- 
grapher and textual critic to propose a cure. Apart from lucky coincidence, what 
lay behind the hodgepodge of nonsense set by the compositor of the first quarto of 
King Lear in m. iv. 118 ff. is probably unattainable : 'swithald footed thrice the old a 
nellthu night more and her nine fold bid her, O light and her troth plight and 
arint thee, with arint thee.' On the special problems involved in the textual criticism 
of Shakespeare's works, see Madeleine Doran, 'An Evaluation of Evidence in 
Shakespearean Textual Criticism', English Institute Annual, 1941 (New York, 1942), 
pp. 95-114, and F. P. Wilson, 'Shakespeare and the "New Bibliography" ', in 
The Bibliographical Society, i8g2-ig42, Studies in Retrospect (London, 1 945) , pp. 1 33-4. 

184 Modern Methods of Textual Criticism 

emendation has been to use it prematurely. Corruptions in the 
Greek and Latin classics (including the New Testament) have 
frequently been assumed without adequate reason — as though, 
indeed, for the mere sake of showing off one's cleverness in pro- 
posing an alternative reading. This 'itch for emending' (pruritus 
emendandi) has resulted in the accumulation of literally thou- 
sands of proposed alterations of passages in the New Testament. 
Those which William Bowyer assembled in the eighteenth cen- 
tury (see p. 116 above) were greatly augmented in the latter 
part of the nineteenth century by a spate of articles and books 
published in Holland by W. C. van Manen, W. H. van der Sande 
Bakhuyzen, D. Harting, S. S. de Koe, H. Franssen, J. M. S. Bal- 
jon, J. H. A. Michelsen, J. Cramer, and others. 1 

In their edition of the Greek New Testament Westcott and 
Hort mark with obeli about sixty passages which they (or one 
of them) suspect involve a 'primitive error', that is, an error older 
than the extant witnesses, for the removal of which one is confined 
to conjectural emendation. 2 According to Schmiedel, 3 the editions 
of Tregelles, Tischendorf, and Weiss contain only one conjecture 

1 For the titles of these works, see Eberhard Nestle's bibliographical list in Urtext 
und Obersetzungen der Bibel (Leipzig, 1897), pp. 55-56 (being a reprint of his article 
'Bibeltext und Bibeliibersetzungen* in Herzog-Hauck's Realencyklopddie fur pro- 
testantische Theologie und Kirche, 3te Aufl.). In the twentieth century the Dutch philo- 
logist, I. I. Hartman, maintained that most errors in the transmission of the classics 
are non-mechanical and therefore inexplicable, and that, therefore, the editor of a 
text is permitted to abandon manuscript evidence whenever in his opinion this is 
demanded by the content; see Hartman's article, 'Ars critica, quid sibi habeat pro- 
positum et qua utatur ratione', in Mnemosyne, N.s., xlviii (ig2o), pp. 227-38. For a 
rebuttal of Hartman's argument, see A. Damste, 'De arte critica', ibid., pp. 424-33, 
and for several discussions of the proper limitations of conjectural emendation, see 
the appendix added to the second edition of Pasquali's Storia delta tradizione e cri- 
tica del testo, entitled 'Congettura e probability diplomatica', pp. 481-6, Paul van 
den Ven, 'Erreurs de methode dans la correction conjecturale des textes byzantins', 
in Byzantion, xxiv (1954), pp. 19-45, an d tne sane remarks by Ludwig Bieler in his 
stimulating essay, 'The Grammarian's Craft', in Folia: Studies in the Christian Per- 
petuation of the Classics, x (1956), pp. 3-42, especially pp. 26 ff. See below, p. 272. 

2 The following are the passages where Westcott and Hort suspected the presence 
of a 'primitive error': Matt. xxi. 28 ff.; xxviii. 7; Mark iv. 28 ; Luke xi. 35 ; John 
iv. I ; vi. 4; viii. g; Acts iv. 25; vii. 46; xii. 25; xiii. 32, 43; xvi. 12; xix. 40; xx. 28; 
xxv. 13; Rom. i. 32; iv. 12; v. 6; vii. 2; xiii. 3; xv. 32; I Cor. xii. 2; 2 Cor. iii. 3, 
17; vii. 8; xii. 7; Gal. v. 1 ; Col.ii. 2, 18, 23; 2 Thess. i. lo; 1 Tim.iv. 3; vi. 7; 2 Tim. 
i. 13; Philem., vs. 9; Heb. iv. 2; x. 1 ; xi. 4, 37; xii. 1 1 ; xiii. 21 ; I Pet. i. 7; iii. 21 ; 
2 Pet. iii. 10, 12; 1 John v. 10; Jude, vss. 1, 5, 22 f.; Rev. i. 20; ii. 12, 13; iii. 1, 
7, 14; ix. 10; xi. 3; xiii. 10, 15, 16; xviii. 12; xix. 13. 

3 Paul W. Schmiedel in Festgabe Adolf Kaegi von Schulern und Freunden dargebracht 
zum 30. September igig (Frauenfeld, 1919), p. 179. 

Modern Methods of Textual Criticism 185 

each: Tregelles at 1 Pet. iii. 7 margin; Teschendorf at Heb. 
xi. 37; and Weiss at Rev. xviii. 14. The apparatus of the twenty- 
fourth edition of Nestle's Greek New Testament includes from 
various sources about 200 conjectures, 90 of which are identified 
by the name of the scholar who first suggested them. 

One must admit the theoretical legitimacy of applying to 
the New Testament a process which has so often been found 
essential in the restoration of the right text in classical authors. 
On the other hand, the amount of evidence for the text of 
the New Testament, whether derived from manuscripts, early 
versions, or patristic quotations, is so much greater than that 
available for any ancient classical author that the necessity of 
resorting to emendation is reduced to the smallest dimensions. 1 
It is perhaps chiefly in the Catholic Epistles and the Apocalypse, 
where the early manuscript evidence is more limited than for any 
other part of the New Testament, that the need for attempting 
conjectural emendation may arise with any degree of urgency. 

1 Of all the emendations proposed for the New Testament perhaps the one which 
has enjoyed the widest favour is the suggestion that the name 'Enoch' has fallen out 
of the text of I Pet. iii. 19, '. . . [the spirit] in which he [Christ or Enoch?] went and 
preached to the spirits in prison'. Since 177a, when Bowyer included this emen- 
dation in the second edition of his Critical Conjectures, down to the present, a large 
number of scholars have given it their approval, including James Moffatt and 
Edgar J. Goodspeed, both of whom introduced it into the text of their translations 
of the New Testament. (For the history of this conjecture and a list of scholars who 
have adopted it, see B. M. Metzger in Journal of Religion, xxxii [1952], pp. 256 f.) 

What shall be said of this conjecture? Admittedly the precise meaning of 1 Pet. 
iii. 19 in the generally received text is difficult to ascertain and a great number of 
interpretations have been proposed, and admittedly the proposed emendation is 
attractive palaeographically (cV <J koj. and 'Evui\ in uncial script are remarkably 
similar : ENU) K& I [ENOJX] ) . Nevertheless, since the introduction of a new subject 
('Enoch') into vs. 19 disturbs an otherwise smooth context and breaks the con- 
tinuity of the argument, the emendation cannot be accepted — for an emendation 
that introduces fresh difficulties stands self-condemned. 


The Causes of Error in the Transmission of the 
Text of the New Testament 

A s the physician must make a correct diagnosis of a disease 
/\ before attempting to effect its cure, so the textual critic 
J. JL must be aware of the several kinds of injuries and dangers 
to which a text transmitted by handwriting is liable to be ex- 
posed before he can rectify the errors. In fact, it is important to 
see not only what might happen, but also what has happened in 
the copying of manuscripts. No systematic attempt is made in 
this chapter to evaluate the relative worth of the variant readings ; 
the purpose is to describe and classify the phenomena rather 
than at this point to prescribe the remedy. 1 


(a) The scribe who was afflicted with astigmatism found it 
difficult to distinguish between Greek letters which resemble 
one another, particularly when the previous copyist had not 

1 For a discussion of the subject from a different point of view, see John W. 
Burgon's The Causes of the Corruption of the Traditional Text of the Holy Gospels, ed. 
by Edward Miller (London, 1896). A suggestive analysis of the origin of errors in 
terms of the several stages involved in the process of copying is given by Eugene 
Vinaver in his 'Principles of Textual Emendation', Studies in French Language and 
Mediaeval Literature Presented to Prof . Mildred K. Pope (Manchester, 1939), pp. 351-69; 
cf. also J. Andrieu, 'Pour l'explication psychologique des fautes de copiste', Revue 
des etudes latines, xxviii (1950), pp. 279-92. For a consideration of the origin of 
transcriptional errors in the copying of the Chinese classics, reference may be made 
to two publications (in Chinese) by Wang Shu-min, professor at the National Uni- 
versity of Taiwan: one is an article on 'Rules of Textual Criticism' in vol. xxiii (2) 
(1952) of the Bulletin of the Institute of History and Philology of Academia Sinica (my in- 
formation comes from the Newsletter of the American Council of Learned Societies, 
v ( 2 ) [ J 954]> PP- 57 f-)> tne oth er is a volume of 422 pages entitled The Textual 
Criticism of the Chinese Classics (Nankang, 1959). According to information kindly 
supplied me by Mr. Andrew T. L. Kuo, in the latter the author analyses 122 ex- 
amples of text-critical problems arising from faulty transcription and deliberate 
alterations. See below, p. 272. 

Causes of Error in Transmission of the Text 187 

written with care. Thus in uncial script the sigma (which was 
customarily made as a lunar sigma) , the epsilon, the theta, and the 
omicron (C, £, @, 0) were sometimes confused. For example, 
in Acts xx. 35 three minuscule manuscripts (614, 161 1, and 
2138) read KomcovTa eSei instead of KcrmwvTas SeT, an error 
which goes back to an uncial ancestor written in scriptio continua. 
In 1 Tim. iii. 16 the earlier manuscripts read OC ('he who') 
while many of the later manuscripts read 6C (the usual con- 
traction for 6e6s, 'God'). The letters gamma, pi, and tau (JT, 
TI, T) were liable to be confused, particularly if the cross-bar 
on the first and last letters were carelessly drawn, or if the right 
leg of the pi were too short. Thus in 2 Pet.ii. 13 some manuscripts 
read c\rc\/7c\JC ('love feasts') and others read c\/7c\Tc\/C 
('deceptions'). If two lambdas were written too close together 
they could be taken as the letter mu, as has happened at Rom. 
vi. 5, where most manuscripts have £kAA\ ('but') but others 
have i\Mt\ ('together'). If a lambda is followed too closely by an 
iota the combination may look like the letter nu (AI and N). 
Thus in the manuscripts of 2 Pet. ii. 18 OAirOiC ('scarcely') 
alternates with ONTOiC ('really'), where the tau and the gamma 
are also confused. Sometimes A and A were mistaken for each 
other, as in Acts xv. 40, where EIHAESd^MENOC ('having 
chosen') appears in codex Bezae as EIIIAE3d\MEN0C ('having 
received 1 ). 

In the generally accepted text of 1 Cor. xii. 13 Paul declares: 
'By one Spirit we were all baptized into one body . . . and all 
were made to drink of one Spirit.' Several witnesses, however, 
conclude the statement thus: '. . . all were made to drink of one 
drink', a variant which arose when scribes misread the letters 
77Mc\ (the usual contraction of the word Trvev/xa) as TIOMZk 
('drink'). Since the word koi was sometimes abbreviated /£, a 
kappa with a heavy dot of ink at the end of the lower diagonal 
line might be taken as the syllable /ecu. This has in fact happened 
at Rom. xii. 11, where the curious variant, 'serving the time' (tu> 
Kaipu> SovAevovres) , arose from the true text, 'serving the Lord' (tw 
Kvpiw SovXevovres) , because a scribe took the contraction of the 
word Kvpiw (KOi) as FiPlti. 

The examples cited above involve the confusion of similar- 
appearing letters in uncial script, used in the production of 
manuscripts down to the ninth century. It is scarcely necessary 

1 88 The Causes of Error in the 

to consider similarities of letters in the subsequent minuscule 
script, for the overwhelming proportion of variant readings 
originated prior to the period of the minuscule manuscripts. 

As was mentioned in the first chapter (see pp. 8-9 above), 
in antiquity non-literary, everyday documents were customarily 
written in a cursive hand in which most of the letters were 
formed without lifting the pen, and abbreviations were freely 
used. Whether any of the books of the Greek Bible ever circu- 
lated in cursive, or semi-cursive, script is an important question 
to which different answers have been given. Wikenhauser, fol- 
lowing Roller, argues that it is unlikely that the original texts 
of the New Testament books were written in cursive script, 
because the rough surface of papyrus made it difficult to use 
that form of writing. 1 On the other hand, Nau pointed out that 
in the books of Chronicles codex Vaticanus contains certain 
permutations of the letters mu, nu, and beta which cannot be 
explained in terms of confusion of uncial script, in which these 
three letters are very different from one another, but which 
are readily explainable in cursive script, where they resemble 
each other closely. For example, the permutation of /x and v 
in 2 Chron. xvi. 7, Avafiei for Avavei; xvii. 8, Mavdavias and Icopav 
for Navdavias and Icopa^i; xxxi. 12-13, Xco^ievias and Maed for 
Xcovevias and Naed; the permutation of |3 and v in 2 Chron. 
xvii. 8, Tojfiahojfieia for Ta>pa8coveia ; the permutation of jS and /x in 
2 Chron. xxi. IO, Aofiva for Aofiva ; xxxvi. 2, AfienaX for AfieiraX. 1 

Another example of a Biblical manuscript which undoubtedly 
goes back to a cursive ancestor is the Berlin fragment of Gene- 
sis, a papyrus copy in semi-cursive script dating from the third 
century a.d. From a study of a wide variety of scribal errors in 
the text, the editors conclude that one or more ancestors were 
written in a typical cursive hand. 3 

1 Alfred Wikenhauser, New Testament Introduction (New York, 1958), p. 67. 

2 See F. Nau in Revue de V orient chrltien, xvi (191 1), pp. 428—9. 

1 See the list of scribal errors collected by Henry A. Sanders and Carl Schmidt, 
The Minor Prophets in the Freer Collection and the Berlin Fragment of Genesis (New York, 
I 9 2 7)> PP- 244-6. See below, p. 272. 

For examples of variant readings in codex Bezae which may have arisen 
in a cursive ancestor, see Paul Glaue's lists in ^eitschr if I fur die neutestamentliche 
Wissenschaft, xlv (1954), pp. 92-94. 

It is possible that p' s had a cursive ancestor, for in Luke xvi. 19 the name of the 
Rich Man is (oVdjiari.) iVeuijs, which is thought to stand by haplography for (ovofiem) 
Nivevrjs. Now, though the similarity of n and vi is not close when written in uncials, 
in some forms of first-century cursive script the syllable ti could have been very 

Transmission of the Text of the New Testament 189 

(b) When two lines in the exemplar from which the scribe was 
making a copy happened to end with the same word or words, 
or even sometimes with the same syllable, his eye might wander 
from the first to the second, accidentally omitting the whole 
passage lying between them. Thus is to be explained the curious 
reading at John xvii. 15 in codex Vaticanus, which lacks the 
words which are enclosed in square brackets: 'I do not pray that 
thou shouldst take them from the [world, but that thou shouldst 
keep them from the] evil one.' In the exemplar from which the 
scribe of this manuscript was copying, the Greek text probably 
stood in the following arrangement : 

avrovs ck rov 


avrovs £k rov 


After the scribe had copied the first line, his eye returned, not to 
the beginning of line 2, but to the beginning of line 4. Such an 
error is called parablepsis (a looking by the side), 1 and is facili- 
tated by homoeoteleuton (a similar ending of lines) . 

Many other examples of omission, called haplography, occur 
in a wide variety of manuscripts. For example, the whole verse 
at Luke x. 32 is lacking in X because the sentence ends with the 
same verb (avTnrapijXdev) as the previous sentence (vs. 31). Codex 
Alexandrinus omits the entire verse at 1 Cor. ix. 2, which ends 
with the same four words (t^ief? eore iv Kvplui) as the previous 
verse. Since the last five words in Luke xiv. 26 and in 27 are 
exactly the same [ov Svvarai elval fxov ^aflr/r?;?) , it is easy to ac- 
count for the accidental omission of verse 27 in more than a 
dozen different manuscripts. The words of 1 John ii. 23, 'He 
who confesses the Son has the Father also', fell out of the later 
manuscripts (on which the King James version depends) because 
of the presence of tov narepa e^ei in adjacent clauses. Other 
interesting cases of error arising from homoeoteleuton are found 

easily confused with the letter v; for examples, see E. M. Thompson, Handbook of 
Greek and Latin Paleography (New York, 1893), chart facing p. 148, and L. Gonzaga 
da Fonseca, Epitome introductionu in palaeographiam graecam (biblicam), ed. altera 
(Rome, 1944), pp. 85, 94 f. 

1 On the possibility of the scribe's eye wandering to a different column of the 
exemplar, see the examples cited by J. Rendel Harris in the American Journal of 
Philology, vi (1885), pp. 25-40. 

1 90 The Causes of Error in the 

in various manuscripts at Luke v. 26; xi. 32; xii. 9; and Rev. 
ix. 2-3. 

Sometimes the eye of the scribe picked up the same word or 
group of words a second time and as a result copied twice what 
should have appeared only once (this kind of error is called 
dittography) . In Acts xix. 34 the cry of the mob, 'Great is 
Artemis of the Ephesians', is given twice in codex Vaticanus. 
Again, instead of the generally accepted text of Acts xxvii. 37, 
'We were in all two hundred and seventy-six (co;) 1 persons in 
the ship (cv t<3 irXoiw)', codex Vaticanus and the Sahidic ver- 
sion read '. . . about seventy-six (<is o?) . . .'. The difference in 
Greek is slight (l7AOla>co? and I7AOI<d<dco?) . 


When scribes made copies from dictation, or even when a 
solitary scribe in his own cell pronounced to himself the words 
which he was transcribing, confusion would sometimes arise 
over words having the same pronunciation as others, but dif- 
fering in spelling (as the English words 'there' and 'their' or 
'grate' and 'great'). During the early centuries of the Christian 
era certain vowels and diphthongs of the Greek language lost 
their distinctive sounds and came to be pronounced alike, as 
they are today in modern Greek. The confusion between a> and 
o was common, accounting for such variants as e^to/iev and 
exo^iev in Rom. v. 1, and c58e and o8e in Luke xvi. 25. 

The diphthong at and the vowel e came to be pronounced alike 
(with a short e sound). As a result the second person plural 
ending -ode sounded the same as the ending of the middle and 
passive infinitive -adai, accounting for the variants ep^eadai and 
epxeade in Luke xiv. 17, ^Xovade and ^Xovadai in Gal. iv. 18, 
and similarly in many other passages. Sometimes the change of 
vowels resulted in an entirely different word. Thus in Matt, xi.16 

1 After the second century B.C. the letters of the Greek alphabet served as 
numerals. In addition to the twenty-four letters of the alphabet, three obsolete 
signs were also employed : the digamma (p) or stigma (?) for 6, the koppa (9 or <; ) for 
90, and the sampi (~^) for 900. The first nine letters of the alphabet stand for units, 
the second nine for tens, the third nine for hundreds : a' = 1 , /)' = 2, y' = 3, S' — 4, 
f' = 5, f' or f' = 6, £' = 7, tj' — 8, 6' = 9, 1 = 10 {ia = 1 1, i/3' = 12, &c), 
k = 20 (m' = 2 1 , &c.) , A' = 30, ft = 40, 1/ = 50, f ' = 60, o' = 70, it' = 80, 9' or 
<;' = 90, p = 100, a' = 200, t' = 300, v' = 400, <f>' — 500, x' = 600, i/r' = 700, 
co' = 800, "Y = 900, ,a =, ,/3 = 2,000, ,y = 3,000, &c. 

Transmission of the Text of the New Testament 191 

erepocs ('others') in some manuscripts varies with eraipois ('com- 
rades') in other manuscripts. 

The pronunciation of ov and of v was sometimes indistinguish- 
able, and accounts for the variation in Rev. i. 5. The translators 
of the King James version followed a text of this verse which had 
Xovaavn ('Unto him that loved us, and washed us from our sins 
in his own blood'), whereas the text used by modern translators 
reads the verb Xvaavri ('. . . and freed us from . . .'), which is 
found in the earlier Greek manuscripts. 

In Koine Greek the vowels 77, 1, and v, the diphthongs ei, 01, 
and vi, and the improper diphthong 77 came to be pronounced 
alike, all of them sounding like ee in English 'feet'. It is not sur- 
prising that one of the commonest kinds of scribal confusion 
involves the substitution of these seven vowels or diphthongs 
for one another. This kind of error, which is commonly called 
itacism, accounts for several extremely odd mistakes present in 
otherwise good manuscripts. For example, in 1 Cor. xv. 54 the 
statement 'Death is swallowed up in victory (vikos)' appears 
in p 46 and B as 'Death is swallowed up in conflict (veiKos)'. 
According to the vision of the Seer on Patmos, around God's 
throne in heaven was 'a rainbow (t/ois) that looked like an 
emerald' (Rev. iv. 3). In X, A, and other witnesses one finds 
the similarly pronounced word 'priests' (Upets) ! 
. In view of the ever-present possibility of committing itacism, 
it is not surprising that the evidence for the Greek personal pro- 
nouns varies widely in New Testament manuscripts (as 77/^1?/ 
vfieis, r)jj.Zvjvfj,lv, and -quas/viias). Problems arise especially in 
the epistolary literature. Did John write his First Letter 'that 
our [r)jMov) joy may be complete' or 'that your (v/xcuv) joy may be 
complete' ? Does Paul include himself with his readers at Gal. 
iv. 28 by using rj/J-eTs, or did he write v/j.eis? Whichever reading 
is judged to be original, it is easy to see how the other arose. In 
the five chapters of 1 Peter the manuscripts contain at least 
seven instances of such an interchange of personal pronouns 
(i. 3, 1 2 ; ii. 2 1 [twice] ; iii. 1 8, 2 1 ; v. 1 o) . Occasionally a confusion 
of personal pronouns took place that produced virtual nonsense 
in the context; Paul's solemn statement in 2 Thess. ii. 14, 'He 
called you (v/nd?) through our gospel, so that you may obtain the 
glory of our Lord Jesus Christ', reads in manuscripts A B D*, 
&c, 'He called us (ij/xa?) through our gospel . . .'. So widespread 

192 The Causes of Error in the 

is this kind of scribal error that the testimony of even the best 
manuscripts respecting personal pronouns is liable to suspicion, 
and one's decision between such variant readings must turn 
upon considerations of fitness in the context. 

Besides eliminating differences in the pronunciation of certain 
vowel sounds, later Greek ceased to give the rough breathing a 
distinctive force. Manuscripts which have rough and smooth 
breathing marks often use them most arbitrarily, so that avrov 
varies with avrov, eo-rqicev (perfect tense of lottj/m) varies with 
eo-TTjKev (imperfect tense of ottjkw), els with els, and other similar 
pairs of words. 1 

In addition to confusion of vowels which sounded alike, cer- 
tain consonants are occasionally interchanged, as in Matt. ii. 6 
ck aov ('from you') becomes e£ oS ('from whom') in X c (cf. 
also Matt. xxi. 19 and Mark xi. 14). In the same category be- 
long instances of confusion between forms of verbs spelled 
with a single or double consonant (e.g. the present and the 
second aorist stems, e/xeAAev and e/xeAev, in John xii. 6) and the 
confusion of different words altogether (e.g. eyewqd-qaav and 
eyeW)9r]oav in John i. 13). Somewhat similar are the readings in 
1 Thess. ii. 7, where the pronunciation of iyevrjOrjuev rjmoi ('we 
were gentle') is almost indistinguishable from eyevT]dr\\xev mjmoi 
('we were babes', see pp. 230-3 below). 

A curious interchange of consonants has taken place at Rev. 
xv. 6, where the description of the seven angels as 'robed in pure 
bright linen' (Aivov) becomes 'robed in pure bright stone' (\i9ov) 
in several early manuscripts (including A C and codices of the 
Vulgate) . At Heb. iv. 1 1 the scribe of codex Claromontanus 
wrote dXrjdelas ('truth') for dneideias ('disobedience'), with quite 
disastrous results to the sense! 


The category of errors of the mind includes those variations 
which seem to have arisen while the copyist was holding a clause 
or a sequence of letters in his (somewhat treacherous) memory 
between the glance at the manuscript to be copied and the 
writing down of what he saw there. In this way one must 

1 For other examples, with references to specific New Testament manuscripts, 
see Blass-Debrunner-Funk, A Greek Grammar of the New Testament and Other Early 
Christian Literature (Chicago, [1961]), § 14. 

Transmission of the Text of the New Testament 193 

account for the origin of a multitude of changes involving the 
substitution of synonyms, variation in the order of words, and 
the transposition of letters. 

(a) The substitution of synonyms may be illustrated by the 
following examples : einev for tyrj ; cV for 0770, and the reverse ; 
evdvs for €i)#€6us, and the reverse ; on for S16V1 ; nepi for imep, and 
the reverse; o/u/zaT6uv for 6<f>daXfi(Lv. 

(b) Variations in the sequence of words is a common pheno- 
menon ; thus the three words iravres ko.1 c/Wti'^ojto in Mark i. 5 
also appear in the order kcu ipa-nTi^ovTo irdvres as well as ko.1 

TraVT€S ef$aTTTl!l,OVTO. 

(c) The transposition of letters within a word sometimes re- 
sults in the formation of a different word, as eXafiov in Mark 
xiv. 65, becomes efiaXov in some manuscripts (and epaWov in 
other manuscripts) . Such alterations of letters sometimes pro- 
duce utter nonsense ; at John v. 39, where Jesus speaks of the 
Scriptures as 'they that bear witness (ai fiaprvpovaai) concerning 
me', the scribe of codex Bezae wrote 'they are sinning (dfiaprd- 
vovaai) concerning me' ! 

(d) The assimilation of the wording of one passage to the slightly 
different wording in a parallel passage, which may have been 
better known to the scribe, accounts for many alterations in the 
Synoptic Gospels. Thus at Matt. xix. 1 7 the reading of the earlier 
manuscripts, 'Why do you ask me about what is good? One there 
is who is good', is changed in later manuscripts to agree with 
the form of Jesus' words as reported in Mark x. 1 7 and Luke 
xviii. 18, 'Why do you call me good? No one is good but God 
alone.' (The King James translators followed the later form of 
text in Matthew.) More than once in the Epistles to the Colossi ans 
and to the Ephesians scribes have introduced into passages of 
one Epistle words and phrases which properly belong to parallel 
passages in the other. Thus, to the statement in Col. i. 14, 'in 
whom we have redemption, the forgiveness of sins', a few later 
Greek manuscripts add the words 'through his blood', a phrase 
derived from the parallel in Eph. i. 7. (Here again the King 
James version follows the secondary form of text.) 


Though perhaps several of the following examples might 
be classified below under the category of deliberate changes 

194 The Causes of Error in the 

introduced for doctrinal reasons, it is possible to regard them as 
unintentional errors committed by well-meaning but sometimes 
stupid or sleepy scribes. 

Words and notes standing in the margin of the older copy 
were occasionally incorporated into the text of the new manu- 
script. Since the margin was used for glosses (that is, synonyms 
of hard words in the text) as well as corrections, it must have 
often been most perplexing to a scribe to decide what to do with 
a marginal note. It was easiest to solve his doubt by putting the 
note into the text which he was copying. Thus it is probable that 
what was originally a marginal comment explaining the moving 
of the water in the pool at Bethesda (John v. 7) was incorporated 
into the text of John v. 3b~4 (see the King James version for the 
addition). Again, it is altogether likely that the clause in later 
manuscripts at Rom. viii. 1, 'who walk not according to the 
flesh but according to the spirit', was originally an explanatory 
note (perhaps derived from vs. 4) defining 'those who are in 
Christ Jesus'. As was mentioned in Chapter I, some manuscripts 
are provided with marginal helps designed to assist the reader 
of the fixed Scripture lessons appointed by the ecclesiastical 
calendar (the Lectionary) . As a result, lectionary formulas, such 
as elnev 6 (cuptos, occasionally crept into the text of certain non- 
lectionary manuscripts (e.g. at Matt. xxv. 31 and Luke vii. 31). 1 
Other errors originated, not because of the exercise of faulty 
judgement, but from the lack of judgement altogether. Only 
heedlessness to a degree that passes comprehension can account 
for some of the absurdities perpetrated by witless scribes. For 
example, after els tovs dylovs of 2 Cor. viii. 4 a good many 
minuscule manuscripts have added the gloss hi^aadai -qixds. It 
appears that a scribe of one of these manuscripts wrote in the 
margin beside Setjaodai rjfjLds the comment ev ttoX\ols riov avn- 
ypd<f>a>v ovtcds evp-qrai ('it is found thus in many of the copies'). 
Then the scribe of a subsequent manuscript (cited by Bengel, 
ad loc.) incorporated this comment on the gloss directly in his 
text as though it were part of the apostle Paul's instructions to 
the Corinthians! 

What is perhaps the most atrocious of all scribal blunders is 

1 For other examples of influence from Iectionaries, reference may be made to 
the present writer's monograph, The Saturday and Sunday Lessons from Luke in the 
Greek Gospel Lectionary (Chicago, 1944), pp. 14-17- 

Transmission of the Text of the New Testament 195 

contained in the fourteenth-century codex 1 09. This manuscript 
of the Four Gospels, now in the British Museum, was transcribed 
from a copy which must have had Luke's genealogy of Jesus 
(hi. 23-38) in two columns of twenty-eight lines to the column. 
Instead of transcribing the text by following the columns in 
succession, the scribe of 109 copied the genealogy by following 
the lines across the two columns. 1 As a result, not only is almost 
everyone made the son of the wrong father, but, because the 
names apparently did not fill the last column of the exemplar, 
the name of God now stands within the list instead of at its close 
(it should end, of course, '. . . Adam, the son of God'). In this 
manuscript God is actually said to have been the son of Aram, 
and the source of the whole race is not God but Phares ! 


Odd though it may seem, scribes who thought were more 
dangerous than those who wished merely to be faithful in copy- 
ing what lay before them. Many of the alterations which may be 
classified as intentional were no doubt introduced in good faith 
by copyists who believed that they were correcting an error or 
infelicity of language which had previously crept into the sacred 
text and needed to be rectified. 3 A later scribe might even re- 
introduce an erroneous reading that had been previously cor- 
rected. For example, in the margin of codex Vaticanus at Heb. 

1 For a description of other manuscripts in which the Lucan genealogy of Jesus is 
confused to a greater or less extent, see Jacob Geerlings, Family U. in Luke (Studies and 
Documents, xxii, Salt Lake City, 1962), pp. 127-37. 

2 For other discussions of this subject, see Eric L. Titus, The Motivation 0/ Changes 
made in the New Testament Text by Justin Martyr and Clement of Alexandria : a Study in the 
Origin of New Testament Variation (Unpublished Diss., University of Chicago, 1942) ; 
C. S. C. Williams, Alterations to the Text of the Synoptic Gospels and Acts (Oxford, 195 1) ; 
Leon E. Wright, Alterations of the Words of Jesus as Quotedin the Literature of the Second 
Century (Cambridge, Massachusetts, 1952); E. W. Saunders, 'Studies in Doctrinal 
Influence on the Byzantine Text of the Gospels', Journal of Biblical Literature, lxxi 
(1952), pp. 85-92 ; K. W. Clark, 'Textual Criticism and Doctrine', Studia Paulina in 
honorem Johannis de Zwaan (Haarlem, 1953), pp. 52-65; Eric Fascher, Textgeschichte 
als hermeneutisches Problem (Halle/S. , 1 953) ; and Manfred Karnetzki, 'Textgeschichte 
als Uberlieferungsgeschichte', £eitschrift fur die neuteslamentliche Wissenschaft, xlvii 
('95 6 )> PP- 1 7°- 180. 

3 Jerome complained of the copyists who 'write down not what they find but 
what they think is the meaning; and while they attempt to rectify the errors of 
others, they merely expose their own' (scribunt non quod inveniunt, sed quod intellegunt; 
et dum alienos errores emendare nituntur, ostendunt suos), Epist. lxxi. 5, Ad Lucinum 
(Migne, P.L. xxii. 671 ; C.S.E.L. lv, pp. 5 f.). For other patristic references to the 
incompetence of some copyists, see pp. 72 and 152 above. 

196 The Causes of Error in the 

i. 3 there is a curiously indignant note by a rather recent scribe 1 
who restored the original reading of the codex, <j>avepc7>v, for 
which a corrector had substituted the usual reading, <j>epcov: 
'Fool and knave, can't you leave the old reading alone and not 
alter it!' (a/xa^eWaTe Kal KaKe, a<j>es tov naAaiov, /xij /xeTa77otet) . 
Andrew of Caesarea in Cappadocia, in his commentary on 
the Book of Revelation, 2 written about the year 600, expressly 
applied the curse recorded in Rev. xxii. 18-19 to those litterati 
who considered that Attic usage 3 and a strictly logical train of 
thought were more worthy of respect and more to be admired 
(agiomoTOTepa Kal ae^ivorepa) than the peculiarities of Biblical 
language. What Andrew refers to is illustrated by an anecdote 
told by Sozomen, a fifth-century lawyer of Constantinople who 
wrote a history of the Church. He relates that at an assembly of 
Cypriot bishops about the year 350, one Triphyllios of Ledra, a 
man of culture and eloquence, was addressing the assembly, and 
in quoting the text, 'Rise, take up your bed and walk', sub- 
stituted the more refined Attic Greek word oki/jlitovs for the 
colloquial Koine word Kpaffiaros ('pallet') used in John v. 8. 
Whereupon a certain Bishop Spyridon sprang up and indig- 
nantly called to him before the whole assembly, 'Are you, then, 
better than he [Jesus] who uttered the word Kpdfifiaros, that you 
are ashamed to use his word?' 4 Despite the vigilance of eccle- 
siastics of Bishop Spyridon's temperament, it is apparent from 
even a casual examination of a critical apparatus that scribes, 
offended by real or imagined errors of spelling, grammar, and 
historical fact, deliberately introduced changes into what they 
were transcribing. 


The Book of Revelation, with its frequent Semitisms and sole- 
cisms, afforded many temptations to style-conscious scribes. It 

1 The scribe was perhaps of the thirteenth century; see below, p. 272. 
1 Josef Schmid, Studien zur Gtschichte dts griechischen Apokalypse-Textes ; i, Der 
Apokalypse-Kommentar des Andreas von Kaisareia (Munich, 1955), p. 262, 11. 3-12. 

3 Cf. Wilhelm Michaelis, 'Der Attizismus und das Neue Testament', £et'tsc/rrt/t 
fur die neutestamentliche Wissenschaft, xxii (1923), pp. 91— 121, and G. D. Kilpatrick, 

'Atticism and the Text of the Greek New Testament', Neutestamentliche Aufsdtze: 
Festschrift fur Prof. Josef Schmid (Regensburg, 1963), pp. 125-37. 

4 Sozomen, Hist. Eccl. i. 1 1 . Eusebius tells us that Tatian ventured to paraphrase 
certain words of the apostle Paul 'as though improving their style' (cuy emSiopdov- 
[icvov avrojv ttjs tf>pdoca>s avvrafciv) , Hist. Eccl. iv. xxxix. 6. 

Transmission of the Text of the Mew Testament 197 

is not difficult to imagine, for example, that the use of the 
nominative case after the preposition d-n-6 (in the stereotyped 
expression, coto d wv koX 6 fy «al 6 ep^ojuevoy, Rev. i. 4) would 
grate on the sensibilities of Greek copyists, and that consequently 
they would insert after cltto either tov, or deov, or Kvpiov in order 
to alleviate the syntax. As a matter of fact, all three of these 
attempts to patch up the grammar are represented today in one 
or more manuscripts. 

The use of Kai joining the finite verb l-rrol-qaev in Rev. i. 6 to 
the participles in vs. 5 strains the rules of Greek concord beyond 
the breaking-point ; scribes mended the syntax by changing the 
indicative to another participle (-n-orfaavTi.) . The genitive case 
of Tre-rrvpcoiAevris in Rev. i. 1 5, which agrees with nothing in its 
clause, was altered by some scribes to the dative and by others 
to the nominative, either of which construes grammatically with 
the rest of the sentence. In Rev. ii. 20 17 Aeyovaa, a pendent 
nominative, was emended to ttjv Xeyovaav, which stands in 
apposition to the immediately preceding words, ttjv yvvaiKa 


Some harmonistic alterations originated unintentionally (ex- 
amples are given in section 1. 3 (d) above) ; others were made 
quite deliberately. Since monks usually knew by heart exten- 
sive portions of the Scriptures (see p. 87 above), the temptation 
to harmonize discordant parallels or quotations would be 
strong in proportion to the degree of the copyist's familiarity 
with other parts of the Bible. The words which belong in John 
xix. 20, 'It was written in Hebrew, in Latin, and in Greek', 
have been introduced into the text of many manuscripts at Luke 
xxiii. 38. The shorter form of the Lord's Prayer in Luke xi. 2-4 
('Father, hallowed be thy name. Thy kingdom come. Give us 
each day our daily bread; and forgive us our sins, for we our- 
selves forgive every one who is indebted to us; and lead us not 
into temptation') was assimilated in many copies of Luke to 
agree with the more familiar, longer form in Matt. vi. 9-13. 
At Acts ix. 5-6 the words spoken to Paul at his conversion are 
conformed in some manuscripts to agree with the parallel ac- 
count in xxvi. 14-15. Frequently Old Testament quotations 
are enlarged from the Old Testament context, or are made to 

826158 O 

i9 8 The Causes of Error in the 

conform more closely to the Septuagint wording. For example, 
the clause in the King James version at Matt. xv. 8, ' [This people] 
draweth nigh unto me with their mouth' — a clause which is not 
found in the earlier manuscripts of Matthew — was introduced 
into later manuscripts by conscientious scribes who compared the 
quotation with the fuller form in the Septuagint of Isa. xxix. 13. 
The earlier manuscripts of John ii. 1 7 quote Ps. lxix. 9 in the 
form 'Zeal for thy house will consume ( /caraway ercu) me'. Since, 
however, the current Septuagint text of this Psalm reads the 
aorist form (/care'^aye), later scribes conformed the Johannine 
quotation to the text of the Septuagint. At Rom. xiii. 9 Paul's 
reference to four of the Ten Commandments is expanded in 
some manuscripts by the addition of another, 'You shall not 
bear false witness.' At Heb. xii. 20 a few witnesses extend the 
quotation from Exod. xix. 13, 'If even a beast touches the 
mountain, it shall be stoned', by adding the words that follow 
in Exodus, 'or thrust through with a dart' (as the King James 
version renders it). 


The work of copyists in the amplifying and rounding off of 
phrases is apparent in many passages. Not a few scribes supposed 
that something is lacking in the statement in Matt. ix. 13, 'For 
I came not to call the righteous, but sinners', and added the 
words 'unto repentance' (from Luke v. 32). So, too, many a 
copyist found it hard to let 'the chief priests' pass without adding 
'the scribes' (e.g. Matt. xxvi. 3), or 'scribes' without 'Pharisees' 
(e.g. Matt, xxvii. 41) ; or to copy out the phrase, 'Your Father 
who sees in secret will reward you' (Matt. vi. 4, 6), without 
adding the word 'openly'. 

Col. i. 23 contains an interesting example illustrating how 
scribes succumbed to the temptation of enhancing the dignity 
of the apostle Paul. In this verse the author warns the Colossians 
against shifting from the hope of the gospel, which 'has been 
preached to every creature under heaven and of which I, Paul, 
became a minister'. The word Sia/coyo?, which means literally 
one who serves, a minister, also came to be used for a lower order 
of the ministry ('deacon'). Perhaps thinking that such a rank 
was less than appropriate for the great apostle to the Gentiles, 

Transmission of the Text of the New Testament 199 

the scribes of X* and P changed Siolkovos to *%>u£ ko.1 airoaroXos, 
while A and Syr h mg read all three nouns ('of which I, Paul, 
became a herald and apostle and minister'). MS. 81 reads 
Sidxovos Kal anooToAos, and the Ethiopic prefers Krjpvtj ko.1 Siol- 
kovos. Here the shorter, less spectacular reading is obviously 

A good example of a growing text is found in Gal. vi. 17, 
where the earliest form of the text is that preserved in p 46 B A C* 
/, 'I bear on my body the marks of Jesus'. Pious scribes could 
not resist the temptation to embroider the simple and unadorned 
'I-qaov with various additions, producing Kvpiov 'I-qoov, as in 
C 3 D° E K L and many other witnesses ; Kvpiov 'Irjoov Xpiarov, in 
Nde Augustine; and Kvpiov 17/xaw 'I-qoov Xptarov, in D^* G 
Syr p Goth Chrysostom, Victorinus, Epiphanius. 


In the earlier manuscripts of Mark i. 2 the composite quo- 
tation from Malachi (iii. 1) and from Isaiah (xl. 3) is intro- 
duced by the formula, 'As it is written in Isaiah the prophet'. 
Later scribes, sensing that this involves a difficulty, replaced ev tu> 
'Hoata rep TTpo<f>-qT7] with the general statement eV tois ■npofyryrais. 
Since the quotation which Matthew (xxvii. 9) attributes to the 
prophet Jeremiah actually comes from Zechariah (xi. 12 f.), it 
is not surprising that some scribes sought to mend the error, 
either by substituting the correct name or by omitting the name 
altogether. A few scribes attempted to harmonize the Johannine 
account of the chronology of the Passion with that in Mark by 
changing 'sixth hour' of John xix. 14 to 'third hour' (which ap- 
pears in Mark xv. 25). At John i. 28 Origen 1 altered BqBavla 
to BrjQafiapa in order to remove what he regarded as a geogra- 
phical difficulty, and this reading is extant today in MSS. 77 W 
33 69 and many others, including those which lie behind the 
King James version. The statement in Mark viii. 3 1 , that 'the 
Son of man must suffer many things . . . and be killed and after 
three days (pera rpeis i^ie/ja?) rise again', seems to involve a 
chronological difficulty, and some copyists changed the phrase 
to the more familiar expression, 'on the third day' (177 rpir-Q -qixipa). 

1 Comm. in Joan. ii. 19 (13) (G.C.S., Origenes, iv. 76. 23-24, ed. Klostermann) . 

200 The Causes of Error in the 

The author of the Epistle to the Hebrews places the golden 
altar of incense in the Holy of Holies (Heb. ix. 4), which is 
contrary to the Old Testament description of the Tabernacle 
(Exod. xxx. 1-6). The scribe of codex Vaticanus and the trans- 
lator of the Ethiopic version correct the account by transfer- 
ring the words to ix. 2, where the furniture of the Holy Place is 


What would a conscientious scribe do when he found that the 
same passage was given differently in two or more manuscripts 
which he had before him? Rather than make a choice between 
them and copy only one of the two variant readings (with the 
attendant possibility of omitting the genuine reading), most 
scribes incorporated both readings in the new copy which they 
were transcribing. This produced what is called a conflation 
of readings, and is characteristic of the later, Byzantine type 
of text. For example, in some early manuscripts the Gospel 
according to Luke closes with the statement that the disciples 
'were continually in the temple blessing God', while others 
read 'were continually in the temple praising God'. Rather than 
discriminate between the two, later scribes decided that it 
was safest to put the two together, and so they invented the 
reading 'were continually in the temple praising and blessing 

In the early manuscripts at Mark xiii. 1 1 Jesus counsels his 
followers not to be 'anxious beforehand' (npofiepifivaiTe) what they 
should say when persecuted. Other manuscripts of Mark read 
'do not practise beforehand' (^po^eAe-rare) , which is the ex- 
pression used also in the Lucan parallel (xxi. 14). Rather than 
choose between these two verbs, a good many copyists of Mark 
gave their readers the benefit of both. In Acts xx. 28 the two 
earlier readings, 'church of God' and 'church of the Lord', are 
conflated in later manuscripts, producing 'the church of the 
Lord and God'. 

Occasionally conflate readings appear even in early manu- 
scripts. For example, codex Vaticanus is alone in reading 
KoXeaavTi /ecu iKavaiaavri at Col. i. 12, whereas all the other 
manuscripts have one or the other participle. 

Transmission of the Text of the New Testament 201 


The number of deliberate alterations made in the interests of 
doctrine is difficult to assess. Irenaeus, Clement of Alexandria, 
Tertullian, Eusebius, and many other Church Fathers accused 
the heretics of corrupting the Scriptures in order to have sup- 
port for their special views. 1 In the mid-second century Marcion 
expunged his copies of the Gospel according to Luke of all ref- 
erences to the Jewish background of Jesus. Tatian's Harmony 
of the Gospels contains several textual alterations which lent 
support to ascetic or encratite views. 

Even within the pale of the Church one party often accused 
another of altering the text of the Scriptures. Ambrosiaster, the 
fourth-century Roman commentator on the Pauline Epistles, 
believed that where the Greek manuscripts differed on any 
important point from the Latin manuscripts which he was ac- 
customed to use, the Greeks 'with their presumptuous frivolity' 
had smuggled in the corrupt reading. In revising the Old Latin 
text of the Gospels, St. Jerome was apprehensive lest he be 
censured for making even slight alterations in the interest of 
accuracy — a fear that events proved to be well founded ! 

The manuscripts of the New Testament preserve traces of two 
kinds of dogmatic alterations : those which involve the elimi- 
nation or alteration of what was regarded as doctrinally un- 
acceptable or inconvenient, and those which introduce into the 
Scriptures 'proof for a favourite theological tenet or practice. 

In transcribing the prologue to the Third Gospel, the scribes 
of several Old Latin manuscripts as well as the Gothic version 

1 See August Bludau, Die Schriftfdlschungen der Hdretiker: ein Beitrag zur Text- 
kritik der Bibel (Miinster/W., 1925). Such changes prove that the autographs of the 
books of the New Testament were no longer in existence, otherwise an appeal would 
have been made directly to them. Their early loss is not surprising, for during per- 
secutions the toll taken by imperial edicts aiming to destroy all copies of the sacred 
books of Christians must have been heavy. Furthermore, simply the ordinary wear 
and tear of the fragile papyrus, on which at least the shorter Epistles of the New 
Testament had been written (see the reference to x^Pttjs in 2 John, vs. 12), would 
account for their early dissolution. It is not difficult to imagine what would happen 
in the course of time to one much-handled manuscript, passing from reader to 
reader, perhaps from church to church (see Col. iv. 16), and suffering damage from 
the fingers of eager if devout readers as well as from climatic changes. (On Peter 
of Alexandria's reference to the original copy of John's Gospel preserved at 
Ephesus [Migne, P.G. xviii. 517], see Juan Leal, 'El autograft) de iv Evangelio 
...', Estndios eclesidstkos, xxxiv [i960], pp. 895-905.) See below, p. 273. 

Transmission of the Text of the New Testament 203 

in the early Church to require rephrasing in order to safeguard 
the virgin birth of Jesus. In ii. 41 and 43 instead of the words 
'his parents' (ol yovets avrov) some manuscripts read 'Joseph 
and Mary'. In ii. 33 and 48 certain witnesses alter the reference 
to Jesus' father either by substituting the name Joseph (as in 
vs. 33) or by omitting it altogether (as in vs. 48). 

In view of the increasing emphasis on asceticism in the early 
Church and the corresponding insistence upon fasting as an 
obligation laid on all Christians, it is not surprising that monks, 
in their work of transcribing manuscripts, should have intro- 
duced several references to fasting, particularly in connexion with 
prayer. This has happened in numerous manuscripts at Mark 
ix. 29, Acts x. 30, and 1 Cor. vii. 5. In Rom. xiv. 17, where the 
kingdom of God is said to be not eating and drinking, 'but 
righteousness and peace and joy in the Holy Spirit', codex 4 in- 
serts after 'righteousness' the words 'and asceticism' (/cat aoKycns) . 
Such interpolations abound in chapter vii of 1 Corinthians. 


In Matt. i. 8 codex Bezae and the Curetonian Syriac insert 
several additional Old Testament names in Jesus' genealogy, 
thereby destroying the Evangelist's intended pattern of fourteen 
generations (i. 17). Besides the instances of agrapha contained 
in certain manuscripts at Luke vi. 4 and Matt. xx. 28 (see 
p. 50 above), there is a curious expansion of Jesus' words to 
Peter in a twelfth- or thirteenth-century minuscule codex of the 
Gospels (no. 713) at Matt. xvii. 26. The passage runs as follows 
(the addition is in italics) : 

Jesus spoke to him, saying, 'What do you think, Simon? From 
whom do kings of the earth take toll or tribute? From their sons or 
from aliens?' And when he said, 'From aliens', Jesus said to him, 
'Then are the sons free?' Simon said, ''Yes? Jesus says to him, ' Then you 
also must give, as being an alien to them. But, not to give offence to them, 
go to the sea and cast a hook, &c.' 

It is noteworthy that this expansion, preserved in a late Greek 
manuscript, was apparently known in the second or third cen- 
tury, for it is witnessed by Ephraem's commentary on Tatian's 
Diatessaron as well as the Arabic form of the Diatessaron. 1 

1 For a discussion of the reading, see J. Rendel Harris, 'The First Tatian Reading 
in the Greek New Testament', Expositor, 8th ser., xxiii (1922), pp. 120-9. 

204 The Causes of Error in the 

Two late minuscule manuscripts of the Book of Acts (614 and 
2147) describe the Philippian jailor as 6 iticjtos Zre</>aro? (Acts 
xvi. 27). Codices 181 and 460 identify the members of 'the 
household of Onesiphorus', to whom the writer of 2 Tim. iv. 19 
sends greetings ; in accord with the apocryphal Acts of Paul and 
Thecla they are said to be 'Lectra, his wife, and Simaeas and 
Zeno, his sons'. 

In the Vulgate at Phil. iv. 3 the words yi^trte ovt,vye ('true 
yokefellow') are appropriately rendered by the Latin germane 
compar. Curiously enough, the Greek text of the bilingual manu- 
scripts F and G make the adjective germane into a proper name, 
reading yv^trie rip^iave avt,vye\ 

The threefold sanctus, d'yto?, ayios, ayio?, sung by the four 
living creatures before the throne of God (Rev. iv. 8), is expanded 
in various manuscripts ; according to Hoskier's collations, one or 
more manuscripts have ayio? four times, six times, seven times, 
eight times (X*), nine times (B and eighty other manuscripts), 
and even thirteen times (MS. 2000). 

According to a scribal addition in the margin of codex S, 
the name of Cleopas' companion on the Emmaus road (Luke 
XXIV. 18) was Simon (o fiera tov KXeuma. iropevofievos Hifwov fy, 
ovx 6 IleTpos dXX 6 erepos). The margin of codex V has the note, 
'The one with Cleopas was Nathanael, as the great Epiphanios 
said in the Panarion. Cleopas was a cousin of the Saviour, the 
second bishop of Jerusalem.' 

A number of interesting expansions appear in manuscripts of 
the early versions. An apocryphal addition in two Old Latin 
manuscripts (a and g 1 ) states that when Jesus 'was baptized, a 
tremendous light flashed forth from the water, so that all who 
were present feared' (Matt. hi. 15). Another Old Latin manu- 
script (k) amplifies Mark's account of the resurrection of Jesus 
by adding at xvi. 3 : 

Suddenly at the third hour of the day 1 there was darkness through- 
out the whole circuit of the land, and angels descended from heaven, 
and he rose 1 in the brightness of the living God, [and] at once they 
ascended with him, and immediately there was light. Then they [the 
women] drew near to the tomb. 

The natural curiosity of readers regarding the identity of 

1 The text at this point is not grammatical; the English renders the general 

Transmission of the Text of the New Testament 205 

persons that are referred to without being named in the New 
Testament prompted scribes to supply proper names. The Sahi- 
dic version gives the name Nineveh to the anonymous Rich 
Man of Luke xvi. 19, as does also the recently discovered p 75 (see 
p. 42 above). The two robbers who were crucified on either 
side of Jesus are variously named in Old Latin manuscripts: 1 

Right-hand Left-hand 

Codex c 

Matt, xxvii. 38 
Mark xv. 27 



Codex / 
Codex r 

Luke xxiii. 32 
Luke xxiii. 32 





The titles of the books of the New Testament were the objects 
of a good deal of elaboration by scribes. It will be obvious that 
titles of, for example, Paul's Epistles were not needed until the 
apostle's correspondence had been collected into one corpus. 
The earliest titles were short and to the point. Later scribes, how- 
ever, were not content with a bare and unadorned title; they 
embroidered it in ways they thought to be in accord with the 
position and reputation of the author. Thus in X and C the 
Book of Revelation is entitled simply A-noKaXvifus "Icodwov. Later 
manuscripts describe John as 'the divine' (i.e. 'the theologian', 
Attok&\v>Pis "Icodwov tov OeoXoyov, MSS. 35, 69, 498, 1957)- 
Others expand by prefixing 'saint' to the name (dyiov 'Icodwov, 
MSS. 1, 2015, 2020, &c), and still others add 'the Evangelist' 
and/or 'the apostle'. The longest and most fulsome title is that 
found in a manuscript at Mount Athos (Hoskier's 236; Greg. 
1775): 'The Revelation of the all-glorious Evangelist, bosom 
friend [of Jesus], virgin, beloved to Christ, John the theologian, 
son of Salome and Zebedee, but adopted son of Mary the Mother 
of God, and Son of Thunder' ('H diroKaXvipis tov iravevo'ogov 
euayyeAioroO, enuj-rr)Qiov <f>l\ov, Trap64vov, ^ycm-r^ne'vou tuj XpiaTOi, 
Icodwov tov OeoXoyov, vlov UaXcop^Tjs xal ZefteSaiov, Qerov Se vlov rfjs 
8€ot6kov Mapias, kcu, vlov fipovTfjs) . The only designation which 
the scribe omits (probably by accident!) is 'apostle'. 

Other scribal additions which eventually found their way 
into the King James version are the subscriptions that are 

1 See J. Rendel Harris, 'On Certain Obscure Names in the New Testament', 
Expositor, 6th ser., i (1900), pp. 161-77, and 'A Further Note on the Names of the 
Two Robbers in the Gospel', ibid., pp. 304-8. 

206 Causes of Error in Transmission of the Text 

appended to the Pauline Epistles, giving information regarding 
the traditional place from which each was sent, as well as in 
some cases what was believed to be the name of the amanuensis 
or of the messenger who was to carry the Epistle. 

Lest the foregoing examples of alterations should give the 
impression that scribes were altogether wilful and capricious in 
transmitting ancient copies of the New Testament, it ought to 
be noted that other evidence points to the careful and pains- 
taking work on the part of many faithful copyists. There are, for 
example, instances of difficult readings which have been trans- 
mitted with scrupulous fidelity. Thus r)\8ev at Gal. ii. 12 yields 
no good sense and can scarcely be the form intended by the 
author. Nevertheless, the scribes of the earliest manuscripts (in- 
cluding p 46 X B D* G al) refrained from correcting it to r^Xdov. 
Another instance of a manifestly erroneous reading is e'C ns 
oTrXayxy - KaL oLKTip/xoi at Phil. ii. i, which could have arisen 
when the original amanuensis misunderstood Paul's pronun- 
ciation of et ti anXdyxva .... However the solecism may have 
originated, the significant point is that all uncials and most minus- 
cules have transmitted it with conscientious exactness. 

Even in incidental details one observes the faithfulness of 
scribes. For example, the scribe of codex Vaticanus copied quite 
mechanically the section numbers which run in one series 
throughout the corpus of the Pauline Epistles, even though this 
series had been drawn up when the Epistle to the Hebrews stood 
between Galatians and Ephesians and is therefore not suitable 
for the present sequence of the Epistles in Vaticanus. 1 These 
examples of dogged fidelity on the part of scribes could be 
multiplied, and serve to counterbalance, to some extent, the 
impression which this chapter may otherwise make upon the 
beginner in New Testament textual criticism. 2 

1 See p. 48 above. 

2 For other examples of the faithfulness of scribes, see B. Blumenkranz, 'Fidelity 
du scribe', Revue du Moyen Age, viii (1952), pp. 323-6, and H. J. Vogels, ' "Librarii 
dormitantes": Aus der Uberlieferung des Ambrosiaster-Kommentars zu den 
Paulinischen Briefen', Sacris Erudiri: Jaarboek ooor Godsdienstwetenschappen, viii (1956), 


The Practice of New Testament 
Textual Criticism 



Perhaps the most basic criterion for the evaluation of 
variant readings is the simple maxim 'choose the reading 
which best explains the origin of the others' . We all fol- 
low this common-sense criterion when confronted with errors 
and 'variant readings' in modern printed books. For example, 
two editions of John Bunyan's classic, The Pilgrim's Progress, di- 
verge in the story of Christian's finding and using a key by which 
he was able to make his escape from Doubting Castle. One 
edition reads 'The lock went desperately hard', while another 
reads 'The lock went damnable hard.' Which is the original 
reading and which has been altered? Did Bunyan write 'desper- 
ately' and a modern editor change it to 'damnable' for some in- 
explicable reason? Or did Bunyan write 'damnable' (using the 
word in its non-profane sense) and someone subsequently 
alter it in order to remove what was deemed to be an offensive 
expression? There can surely be no doubt what the answer is. 1 
Another criterion which we instinctively recognize to be basic 
is that the reconstruction of the history of a variant reading is 
prerequisite to forming a judgement about it. For example, in 
the earlier printings of the second edition of the unabridged 
Webster's New International Dictionary of the English Language 
(Springfield, 1934) there stands the entry: 

dord (dord), n. Physics & Chem. Density. 
Now, it is a fact that there is no English word 'dord' ; its presence 

1 The example given above has been simplified ; at least three different modifi- 
cations of the original reading have been introduced by editors or printers. Besides 
'desperately hard' the present writer has seen copies which read 'extremely hard' 
and 'very hard'; for the authentic text, reference may be made to James B. 
Wharey's collation of the first eleven editions of The Pilgrim's Progress (Oxford, 
1928), all of which read 'damnable hard'. 

208 The Practice of New Testament Textual Criticism 

in this venerable dictionary is the result of what may be called 
an accidental 'scribal error'. As was acknowledged later by the 
publishers, the entry originated in the confusion of the abbrevia- 
tion, given both as a lower-case letter and an upper-case letter, 
of the word 'density', and was intended to stand thus: 

d. or D., Physics & Chem. Density. 

Not noticing the periods, someone took the collocation of letters 
as a word and called it a noun. The remarkable thing is that the 
error escaped detection for more than a decade, during which 
the volume was reprinted several times. 

Another example of a clerical mistake, this one occurring 
in the highly esteemed Who's Who in America, arose because of 
incompetent judgement. The first time that the biography 
of Thomas Mann appeared in this distinguished cyclopedia of 
famous persons, he was given, quite gratuitously, a middle name. 
In the volume for 1939 the entry reads, in bold-face type, 'Mann, 
Thomas Schriftst' ; in subsequent volumes, however, 'Schriftst' is 
lacking. Which form of the name is correct? An examination of 
the volume Wer Isfs, which is the German counterpart of Who 's 
Who, discloses that 'Schriftst.' is the customary abbreviation of 
the German word for 'author' (Schriftsteller) . Obviously someone 
who prepared the biographical sketch for the American volume 
mistakenly took the abbreviation of Mann's occupation to be 
his middle name. 

The two criteria mentioned earlier are capable of very wide 
application, and include by implication a great many other sub- 
sidiary criteria. It will be useful, however, to specify in more 
precise detail the various considerations which scholars take 
into account in evaluating variant readings of New Testa- 
ment witnesses. It is usual to classify these criteria in terms 
of (1) External Evidence and (2) Internal Evidence; the latter 
involves what Hort termed Transcriptional Probabilities and 
Intrinsic Probabilities. (Here the student should re-read the 
account of the principles underlying Westcott and Hort's 
edition, pp. 129-35 above, as well as the summary of B. H. 
Streeter's subsequent contributions to textual theory, pp. 169-73 
above.) The following is a list of the chief considerations which 
the textual critic takes into account when evaluating variant 
readings in the New Testament. 

The Practice of New Testament Textual Criticism 209 

An outline of basic criteria and considerations to be taken into account 
in evaluating variant readings 

I. external evidence, involving considerations bearing upon: 

( 1) The date of the witness. (Of even greater importance than 
the age of the document itself is the date of the type of text 
which it embodies. The evidence of some minuscule manu- 
scripts (e.g. 33, 81, and 1739) is of greater value than that of 
some of the later or secondary uncials.) 

(2) The geographical distribution of the witnesses that agree 
in supporting a variant. (One must be certain, however, that 
geographically remote witnesses are really independent of one 
another. Agreements, for example, between Old Latin and Old 
Syriac witnesses may be due to influence from Tatian's Dia- 

(3) The genealogical relationship of texts and families of 
witnesses. (Witnesses are to be weighed rather than counted. 
Furthermore, since the relative weight of the several kinds of 
evidence differs for different kinds of variants, there can be no 
merely mechanical evaluation of the evidence.) 

II. internal evidence, involving two kinds of probabilities : 

A. Transcriptional Probabilities depend upon considerations of palaeo- 
graphical details and the habits of scribes. Thus : 

(1) In general the more difficult reading is to be preferred, 
particularly when the sense appears on the surface to be erro- 
neous, but on more mature consideration proves itself to be 
correct. (Here 'more difficult' means 'more difficult to the 
scribe', who would be tempted to make an emendation. The 
characteristic of most scribal emendations is their superficiality, 
often combining 'the appearance of improvement with the 
absence of its reality' [Westcott-Hort, ii, p. 27]. Obviously the 
category 'more difficult reading' is relative, and a point is some- 
times reached when a reading must be judged to be so difficult 
that it can have arisen only by accident in transcription.) 

(2) In general the shorter reading is to be preferred, except 
where (a) parablepsis arising from homoeoteleuton may have 
occurred; or where (b) the scribe may have omitted material 
which he deemed to be (i) superfluous, (ii) harsh, or (iii) con- 
trary to pious belief, liturgical usage, or ascetical practice. 

2io The Practice of New Testament Textual Criticism 

(Compare Griesbach's fuller statement of this criterion, p. 120 

(3) Since scribes would frequently bring divergent passages 
into harmony with one another, in parallel passages (whether 
involving quotations from the Old Testament or different ac- 
counts of the same event or narrative) that reading is to be 
preferred which stands in verbal dissidence with the other. 

(4) Scribes would sometimes (a) replace an unfamiliar word 
with a more familiar synonym, (b) alter a less refined gram- 
matical form or less elegant lexical expression in accord with 
Atticizing preferences, or (c) add pronouns, conjunctions, and 
expletives to make a smooth text. 

B. Intrinsic Probabilities depend upon considerations of what the 
author was more likely to have written, taking into account : 

(1) the style and vocabulary of the author throughout the 

(2) the immediate context, 

(3) harmony with the usage of the author elsewhere, and, ,in 
the Gospels, 

(4) the Aramaic background of the teaching of Jesus, 

(5) the priority of the Gospel according to Mark, and 

(6) the influence of the Christian community upon the for- 
mulation and transmission of the passage in question. 

Not all of these criteria are applicable in every case. The 
critic must know when it is appropriate to give primary con- 
sideration to one type of evidence and not to another. Since 
textual criticism is an art as well as a science, it is understand- 
able that in some cases different scholars will come to different 
evaluations of the significance of the evidence. This divergence is 
almost inevitable when, as sometimes happens, the evidence is so 
divided that, for example, the more difficult reading is found 
only in the later witnesses, or the longer reading is found only in 
the earlier witnesses. 

One of the perennial dangers which confront scholars in every 
discipline is the tendency to become one-sided and to over- 
simplify their analysis and resolution of quite disparate ques- 
tions. In textual criticism this tendency can be observed when 
a scholar, becoming enamoured of a single method or criterion 

The Practice of New Testament Textual Criticism 2 1 1 

of textual analysis, applies it more or less indiscriminately to a 
wide variety of problems. For example, at the beginning of the 
twentieth century Adalbert Merx devoted three learned volumes 
to the attempt to prove that the Western text is closest to the 
original, and that its best representative is the Sinaitic Syriac 
palimpsest. 1 Nearly half a century ago Adolf von Harnack, be- 
lieving that the principles of New Testament criticism needed 
to be revised, suggested that the Latin Vulgate had been largely 
overlooked in the arsenal of the critic's tools. 2 Though it may 
well be that some scholars had not given due consideration 
to Jerome's contributions to textual criticism, Harnack's pro- 
posal to accord the Vulgate text a preponderant weight in the 
evaluation of variant readings found many adverse critics, 
including several Roman Catholic scholars. 3 Similarly, von 
Soden's extravagant estimate of the influence which Marcion 
and Tatian exerted upon the text of the New Testament and 
A. C. Clark's repeated appeal to the longer text are generally 
regarded today as warnings against a one-sided and un- 
warranted over-simplification of the evidence. 4 


To teach another how to become a textual critic is like teach- 
ing another how to become a poet. The fundamental principles 
and criteria can be set forth and certain processes can be de- 
scribed, but the appropriate application of these in individual 

1 Adalbert Merx, Die vier kanonischen Evangelien nach ihrem altesten bekannttn 
Texte . . .; 2ter Theil, Erlduterungen (Berlin, igo2, 1905, 191 1). 

2 Adolf von Harnack, Z ur Revision der Prinzipien der neutestamentlichen Textkritik: 
die Bedeutung der Vulgatafiir den Text der katholischen Briefe und der Anteil des Hierony- 
mus an dem Vbersetzungswerk ( = Beitrdge zur Einleitung in das JVeue Testament, 7. Teil; 
Leipzig, 1916), and 'Studien zur Vulgata des Hebraerbriefs' in Sitzungsberichte der 
Preussischen Akademie, 1920, pp. 179-201 (= Studien zur Geschichte des Neuen Testa- 
ments und der Allen Kirche, i, ed. by Hans Lietzmann [Berlin, 1931], pp. igi-234). 

3 See, for example, Michael Hetzenauer, De recognitione principiorum criticae textits 
Novi Testamenti secundum Adolfum de Harnack (Ratisbonae, 1921; reprinted from 
Lateranum, 1920, no. 2), and J. Belser, 'Zur Textkritik der Schriften des Johannes', 
Theologische Quartalschrift, xcviii (1916), pp. 145-84. 

4 Housman poked fun at this type of textual critic. 'We must have', he wrote, 
'no favourite method. An emendator with one method is as foolish a sight as a 
doctor with one drug. The scribes knew and cared no more about us and our 
tastes than diseases care about the tastes of doctors; they made mistakes not of one 
sort but of all sorts, and the remedies must be of all sorts too' (A. E. Housman, ed., 
M. Manilii Astronomicon, liber primus [Cambridge, 1903], pp. liiif.). 

aia The Practice of New Testament Textual Criticism 

cases rests upon the student's own sagacity and insight. With 
this caveat in mind, the beginner will know how to estimate the 
following simplified description of text-critical methodology. 

As a preliminary step in analysing and evaluating the evidence 
found in a critical apparatus, the several variant readings 
should be set down in a list, each with its supporting witnesses. 
This will help one to see clearly the point at issue, and whether 
the documents have two principal readings or more. 

In the evaluation of the evidence the student should begin 
with external considerations, asking himself which reading is sup- 
ported by the earliest manuscripts and by the earliest type of text. 
Readings which are early and are supported by witnesses from 
a wide geographical area have a certain initial presumption in 
their favour. On the other hand, readings which are supported 
by only Koine or Byzantine witnesses (Hort's Syrian group) may 
be set aside as almost certainly secondary. 1 The reason that jus- 
tifies one in discarding the Koine type of text is that it is based 
on the recension prepared near the close of the third century 
by Lucian of Antioch, or some of his associates, who deliber- 
ately combined elements from earlier types of text. Despite the 
fact that it appears in a large majority of Greek manuscripts (for 
it was adopted, with subsequent modifications, as the received 
text of the Greek Orthodox Church), the abundance of witnesses 
numerically counts for nothing in view of the secondary origin 
of the text- type as a whole. 

To facilitate the process of ascertaining which types of text 
support the several variant readings, the student should become 
thoroughly familiar with the following tables of witnesses. One 
must beware, however, of supposing that these text-types are 
static and exactly defined entities ; on the contrary, each text-type 
involves a process 2 of textual development which, though dis- 
tinctive and characteristic as a whole, cannot be isolated within 
precisely determined boundaries. 

1 Theoretically it is possible that the Koine text may preserve an early reading 
which was lost from the other types of text, but such instances are extremely rare 
(one example is discussed on pp. 238-9 below). For a survey of previous evalua- 
tions of such readings, see the chapter on the Lucianic Recension in B. M. Metzger, 
Chapters in the History of New Testament Textual Criticism (Leiden and Grand Rapids, 
1963), pp. 1-41. See below, p. 273. 

1 Cf. E. C. ColwelPs discussion of 'The Origin of Texttypes of New Testament 
Manuscripts', Early Christian Origins: Studies in honor of Harold R. Willoughby, ed. by 
Allen Wikgren (Chicago, 1961), pp. 128-38, especially pp. i36f. 

The Practice of New Testament Textual Criticism 213 


Gospels: AEFGHKPSVW (in Matt, and Luke viii. 13- 
xxiv. 53) IIW (in Luke and John) Q and most minuscules. 
Acts : H a L ap P a 049 and most minuscules. 
Epistles : L ap 049 and most minuscules. 
Revelation: 046 051 052 and many minuscules. 1 


The forms of text which antedate the Koine or Byzantine text 
include the Western group of texts, the so-called Caesarean text, 
and the Alexandrian (Hort's 'Neutral') text. 2 

The Western group of texts 

Though some have held that the Western text was the de- 
liberate creation of an individual or several individuals who 
revised an earlier text, 3 most scholars do not find this type of 
text homogeneous enough to be called a textual recension; it 
is usually considered to be the result of an undisciplined and 
'wild' growth of manuscript tradition and translational activity. 

The Western type of text can be traced back to a very early 

date, for it was used by Marcion (and probably Tatian), 

Irenaeus, Tertullian, and Cyprian. Its most important witnesses 

are codex Bezae and the Old Latin manuscripts, all of which 

are characterized by longer or shorter additions and by certain 

striking omissions. So-called 'Western' texts of the Gospels, 

Acts, and Pauline Epistles circulated widely, 4 not only in North 

* The Byzantine text of the Book of Revelation is less homogeneous than it is in 
other books of the New Testament, for the Greek Orthodox Church has never in- 
cluded readings from the Apocalypse in its lectionary system — a system which 
exerted a stabilizing influence on the Byzantine text of other books of the New 

2 For fuller descriptions of each of these three pre-Koine types of text, as well as 
lists of witnesses that support each type in the several natural divisions of the New 
Testament (Gospels, Acts, Pauline Epistles, Catholic Epistles, and Revelation), 
see M.-J. Lagrange, Critique textuelle; ii, La Critique rationnelle (Paris, 1935). Some- 
what different lists are given in the preface of August Merk's Novum Testamentum 
graece el latine, 8th ed. (Rome, 1957). 

3 So, for example, James Hardy Ropes, The Text of Acts ( = The Beginnings of 
Christianity, Part I, vol. iii; London, 1926), pp. viii ff., and W. H. P. Hatch, The 
' 'Western' Text of the Gospels (Evanston, 1937)- Ropes suggests that the preparation 
of the Western text was involved in the work of forming the primitive canon of the 
New Testament. 

4 The Catholic Epistles and the Book of Revelation seem not to have existed in a 
characteristically Western form of text. 

826158 P 

214 The Practice of New Testament Textual Criticism 

Africa, Italy, and Gaul (which are geographically 'Western'), 
but also in Egypt 1 and (in somewhat different text-forms) in the 
East. These latter text-forms are represented by the Sinaitic and 
Curetonian manuscripts of the Old Syriac, by many of the 
marginal notes in the Harclean Syriac, and perhaps by the 
Palestinian Syriac. 

Westcott and Hort regarded the Western text as almost 
totally corrupt and accepted as original in it only what they 
called 'Western non-interpolations'. As was mentioned above, 
subsequent scholars (e.g. Merx and A. C. Clark) reacted 
against this one-sided view with an equally one-sided preference 
for the Western text. Today such extreme positions for and 
against Western forms of text find little favour, for most textual 
scholars recognize that all of the pre-Koine forms of text deserve 
a hearing, and that any one of them may preserve original 
readings which have been lost to the other text-types. 


Gospels: D W (in Mark i. i-v. 30) 0171, the Old Latin; Syr 3 
and Syr c (in part) , early Latin Fathers, Tatian's Diatessaron. 

Acts: p 2 «p38p48 D 383 614 Syr h m «, early Latin Fathers, the 
Commentary of Ephraem (preserved in Armenian) . 

Pauline Epistles : the Greek-Latin bilinguals Dp Ep Fp Gp ; Greek 
Fathers to the end of the third century ; the Old Latin and 
early Latin Fathers; Syrian Fathers to about a.d. 450. 

The Caesarean text and its witnesses 
B. H. Streeter identified the text that Origen used at Caesarea 
and associated it with the text in 6>, fam. 1, fam.13, and other 
witnesses. Subsequent investigations by Lake, Blake, and New 
showed that the Caesarean text probably originated in Egypt 
and was brought by Origen to Caesarea, from where it was 
carried to Jerusalem (a number of Caesarean witnesses con- 
tain the so-called 'Jerusalem colophon' ; see the description of 

1 Evidence for the presence of the Western text in Egypt is found chiefly in 
several papyri (e.g. p z °, p 38 , and p +8 ). The common opinion that Clement of Alexan- 
dria was accustomed to use a Western form of text (a view based on P. M. Bar- 
nard, The Biblical Text of Clement of Alexandria [ Texts and Studies, v. 5 ; Cambridge, 
1899]) must now be modified in the light of further research by R.J. Swanson (The 
Gospel Text of Clement of Alexandria [Unpublished Diss., Yale University, 1956]), who 
finds that in the Stromala Clement's quotations of Matthew and John are twice as 
often from the Egyptian (= Alexandrian) text as from the Western text. 

The Practice of New Testament Textual Criticism 215 

codex 157 in Chapter II above), to the Armenians (who had 
a colony in Jerusalem at a very early date) , and thence to the 
Georgians (codex Koridethi belongs to Georgia). 

The special character of the Caesarean text is its distinctive 
mixture of Western readings and Alexandrian readings. Accord- 
ing to Lagrange, evidently its maker knew both and made a 
kind of compromise ; in substance he followed the Alexandrian 
text while retaining any Western readings which did not seem 
too improbable, for the latter text was widely current, although 
the former was the better. One may also observe a certain 
striving after elegance, and thus consideration for the needs of 
the Church. 1 

According to more recent investigations made by Ayuso and 
others, 2 it is necessary to distinguish two stages in the devel- 
opment of the Caesarean text (at least for Mark). The Old 
Egyptian text which Origen brought with him to Caesarea may 
be called the pre-Caesarean text. This is preserved in p«, W 
(in Mark v. 31-xvi. 20), fam. 1, fam. 13, 28, and many Greek 
lectionaries. At Caesarea and in its subsequent development, the 
pre-Caesarean text took on the form to which we are led back 
by the common evidence of @, 565, and 700, many of the 
citations of Origen and Eusebius, and the Old Armenian and 
Old Georgian versions (this form is the Caesarean text proper) . 
There also seems to be some degree of affinity between the Old 
Syriac (Syr s - c ) and the Caesarean text. In short, the Caesarean 
text appears to be the most mixed and the least homogeneous 
of any of the groups which can be classified as distinct text- 

The Alexandrian text 
It is widely agreed that the Alexandrian text was prepared by 
skilful editors, trained in the scholarly traditions of Alexandria. 3 
The text on which they relied must have already been an ancient 
text in all important points. Until recently the two chief wit- 
nesses to this form of text were B and X, dating from about the 

1 M.-J. Lagrange, La Critique textuelle; ii, La Critique rationnelle (Paris, 1935), 
pp. 163 ff. 

1 For a survey of the history of the investigation of the Caesarean text, see 
pp. 42-71 of the volume mentioned in n. 1 on p. 212 above. See below, p. 273. 

3 See G. Zuntz, The Text of the Epistles, a Disquisition upon the Corpus Paulinum 
(London, 1953), pp. 272-6. 

2i6 The Practice of New Testament Textual Criticism 

middle of the fourth century. With the discovery, however, of 
p 66 and p 75 , both dating from about the end of the second or 
the beginning of the third century, 1 proof is now available that 
Hort's 'Neutral' text goes back to an archetype which must be 
put early in the second century. 2 This earlier form of the Alex- 
andrian text, which may be called the proto-Alexandrian text, 
is generally shorter than the text presented in any of the other 
forms, the Western being the longest. Furthermore, the proto- 
Alexandrian text appears not to have undergone the systematic 
grammatical and stylistic polishing that was given to other 
texts, including the later form of the Alexandrian text itself. 

Though most scholars have abandoned Hort's optimistic view 
that codex Vaticanus (B) contains the original text almost un- 
changed except for slips of the pen, they are still inclined to 
regard the Alexandrian text as on the whole the best ancient 
recension and the one most nearly approximating the original. 


(i) Proto-Alexandrian: 

p« (in Acts) p 46 p 66 p 75 X B Sahidic (in part), Clement of 
Alexandria, Origen (in part), and most of the papyrus 
fragments with Pauline text. 

(2) Later Alexandrian : 

(C) L T W (in Luke i. i-viii. 12 and John) (X) Z A (in 
Mark) 3 W (in Mark; partially in Luke and John) 33 
57g 892 1 241 Bohairic. 

Acts : p 50 A (C) ¥ 33 81 104 326. 

Pauline Epistles: A (C) fD> I ¥ 33 81 104 326 1739. 

Catholic Epistles: p 20 p 23 A (C) W 33 81 104 326 1739. 

Revelation: A (C) 1006 161 1 1854 2053 2344; less good p 47 ^ 

After having ascertained the text-types represented by the 
evidence supporting each of the variant readings under examina- 
tion, the student should draw a tentative conclusion as to the 
preferred reading on the basis of considerations bearing on the 
age of the manuscripts, the geographical spread of the witnesses 

1 Herbert Hunger, however, dates p 66 earlier in the second century; see n. i on 
p. 40 above. z See below, p. 273. 

The Practice of New Testament Textual Criticism 217 

which join in support of a given reading, and the textual types 
to which they belong. Due appreciation of the implications of 
genealogical relationship among manuscripts prevents one from 
favouring a reading merely because a large number of witnesses 
may support it. 

The next step in the process of evaluating variant readings is 
to appeal to internal evidence, beginning with transcriptional 
probabilities. Which reading is the more difficult — that is, more 
difficult to the scribe? Other things being equal, the reading 
which puzzled the scribe is most likely to be correct. On the 
other hand, there is a point at which what is relatively difficult 
becomes absolutely difficult, and therefore impossible to be re- 
garded as original. 

Some readings were favoured by scribes because they sup- 
ported current beliefs and practices in their part of the Chris- 
tian world. Hence the textual critic will need to have the fullest 
knowledge of the development of Christian doctrine and cultus, 
as well as all the heretical aberrations in the early Church. It goes 
without saying that acquaintance with palaeographical features 
of uncial and minuscule hands, along with a knowledge of dia- 
lectical variations in Greek orthography and syntax, will often 
suggest the correct evaluation of a variant reading. When deal- 
ing with a passage in the Synoptic Gospels it is necessary to 
examine the evidence of parallel passages. The harmonization 
of the Evangelists is by definition a secondary procedure ; there- 
fore the supreme rule for editors of the text is to give each 
Gospel its own proper character. This means that ordinarily 
the reading which differs from a parallel passage (particularly 
when the evidence for the reading of the parallel is firm) should 
be preferred. Likewise, in instances of quotations from the Old 
Testament, the text and apparatus of the Septuagint must be 
consulted. Since scribes tended to make New Testament quo- 
tations conform to the text of the Septuagint, readings which 
diverge from the Old Testament should not be rejected without 
the most careful consideration. 

Finally, the student may appeal to intrinsic probability. The 
reading deemed original should be in harmony with the author's 
style and usage elsewhere. Since, however, it is conceivable that 
several variant readings may fulfil this requirement, the textual 
critic should be guided more by negative judgements delivered 

218 The Practice of Mew Testament Textual Criticism 

by intrinsic evidence than by positive judgements. The appro- 
priate question to ask is whether intrinsic evidence opposes 
the conclusion commended by genealogical considerations, the 
geographical distribution of witnesses, and transcriptional prob- 

Sometimes it happens that the only reading which seems to be 
in harmony with the author's usage elsewhere is supported by 
the poorest external evidence. In such cases the decision of the 
textual critic will be made in accord with his general philosophy 
of textual methodology. It is probably safest for the beginner to 
rely upon the weight of external evidence rather than upon what 
may be an imperfect knowledge of the author's usage. 

In the course of time the student will observe that generally 
the reading which is supported by a combination of Alexan- 
drian and Western witnesses is superior to any other reading. 
There is, however, an exception to this observation; in the 
Pauline Epistles the combination of B D G is ordinarily not of 
great weight. The reason for this is that though B is purely 
Alexandrian in the Gospels, in the Pauline Epistles it has a 
certain Western element. Hence the combination of B plus one 
or more Western witnesses in Paul may mean only the addition 
of one Western witness to others of the same class. 

The combination of Western and Caesarean witnesses does 
not usually possess exceptional weight, for the Caesarean text 
was probably formed from a base which had Western affilia- 

In the evaluation of readings which are supported by only 
one class of witnesses, the student will probably find that true 
readings survive frequently in the Alexandrian text alone, less 
frequently in the Western group alone, and very rarely only in 
Caesarean witnesses. As a rule of thumb, the beginner may 
ordinarily follow the Alexandrian text except in the case of 
readings contrary to the criteria which are responsible for its 
being given preference in general. Such a procedure, however, 
must not be allowed to degenerate into merely looking for the 
reading which is supported by B and K (or even by B alone, as 
Hort was accused of doing) ; in every instance a full and careful 
evaluation is to be made of all the variant readings in the light of 
both transcriptional and intrinsic probabilities. The possibility 
must always be kept open that the original reading has been 

The Practice of New Testament Textual Criticism 219 

preserved alone in any one group of manuscripts, even, in ex- 
tremely rare instances, in the Koine or Byzantine text. 

It remains now to put into practice these principles. Lest, 
however, the student imagine that the procedures of criticism 
are stereotyped and doctrinaire, this section may be concluded 
on a lighter vein with a quotation from a scintillating essay on 
textual criticism by A. E. Housman: 

Textual criticism is not a branch of mathematics, nor indeed an 
exact science at all. It deals with a matter not rigid and constant, 
like lines and numbers, but fluid and variable; namely the frailties 
and aberrations of the human mind, and of its insubordinate servants, 
the human fingers. It is therefore not susceptible of hard-and-fast rules. 
It would be much easier if it were; and that is why people try to pre- 
tend that it is, or at least behave as if they thought so. Of course you 
can have hard-and-fast rules if you like, but then you will have false 
rules, and they will lead you wrong; because their simplicity will 
render them inapplicable to problems which are not simple, but 
complicated by the play of personality. A textual critic engaged 
upon his business is not at all like Newton investigating the motions 
of the planets : he is much more like a dog hunting for fleas. If a dog 
hunted for fleas on mathematical principles, basing his researches on 
statistics of area and population, he would never catch a flea except 
by accident. They require to be treated as individuals; and every 
problem which presents itself to the textual critic must be regarded 
as possibly unique. 1 


The following passages have been chosen in order to provide 
illustrative examples of various kinds of text-critical problems. 
To prevent monotony in the exposition and to emphasize that 
no one stereotyped method of textual analysis is suited to all 
problems, the presentation of the kinds and nature of the 
evidence will be varied in sequence and in development of 
argument. The discussion begins with relatively simple prob- 
lems, for which one can usually find clear and unambiguous 
solutions, and concludes with more complex problems, where 
the probabilities are much more evenly divided and where the 

1 A. E. Housman, 'The Application of Thought to Textual Criticism', Proceed- 
ings of the Classical Association, August, 1921, xviii (London, 1922), pp. 68-69. (This 
essay is now more widely available in John Carter's edition of several of Hous- 
man's critical writings, entitled Selected Prose [Cambridge, 1961], pp. 131-50.) 

220 The Practice of Mew Testament Textual Criticism 

critic must sometimes be content with choosing the least un- 
satisfactory reading, or even with admitting that he has no 
clear basis for choice at all. 

It is customary in a critical apparatus to use abbreviations of 
certain Latin words as a concise and 'international' working 
language. The following are in general use: 

pc (pauci) = a few other manuscripts 

al (alii) = other manuscripts 

pm (permulti) = very many other manuscripts 

pi (plerique) = most other manuscripts 

rell (reliqui) = the remaining witnesses 

vid (videtur) = as it seems, apparently 

omn (omnes) = all manuscripts 

codd (codices) — manuscripts of a version or Church Father as 
distinguished from the edition 

ap (apud) = in the writings of, on the authority of (e.g. 

Papias ap Eusebius) 

p' (partim) = divided evidence (e.g. OrigP' signifies that 

Origen is inconsistent in his quotations of the 
same passage) 

2/4 = divided evidence (e.g. Orig 2/4 signifies that 

in two cases out of four quotations of the 
same passage Origen supports a given read- 

An asterisk (*) placed after the siglum of a manuscript in- 
dicates that the manuscript at the passage referred to has been 
corrected and that the original reading is being cited ; a superior 
letter ( c ) placed after the siglum indicates that the corrected 
reading is being cited. Sometimes the work of more than one 
corrector can be differentiated (for the several ways in which 
such information is cited, see the descriptions of codex Sinai ticus, 
p. 46 above, and codex Ephraemi, p. 49 above). When the 
siglum of a manuscript is enclosed within parentheses, this 
signifies that the manuscript supports the chief point of the 
variant reading but differs in minor respects. 

In this connexion a warning may not be out of place : in some 
apparatus critici of the New Testament the sigla of uncial manu- 
scripts are often cited without the superior letter (or the inferior 
numeral) that serves to distinguish certain manuscripts from 
others designated by the same siglum. Thus D simpliciter often 

The Practice of New Testament Textual Criticism 221 

stands for codex Claromontanus (instead of Dp [or D 2 ]) as well 
as for codex Bezae. In such instances one must be alert to dis- 
tinguish between the two manuscripts by observing whether the 
variant reading occurs in the Gospels or Acts (in this case D = 
codex Bezae), or whether the variant occurs in the Pauline 
Epistles (in this case D = codex Claromontanus). In the fol- 
lowing textual analyses the witness's siglum with the superior 
letter (when this is appropriate) will be used in accord with the 
descriptions of manuscripts given above in Chapter II : that is to 
say, the superior letter a after a siglum indicates a manuscript 
that contains the Acts and the Catholic Epistles, p indicates 
the Pauline Epistles, and r the Book of Revelation. 

In the King James version of Acts vi. 8 Stephen is described as 
'full of faith and power' (nX^p-qs marews ko.1 8wd/i£wj), whereas in 
the Revised Standard Version and the New English Bible he is 
said to be 'full of grace and power' (nX-qp-qs xdpiros nal Swa/iew?). 
The difference in the English versions represents not variant 
renderings of the same Greek word but variant readings in the 
basic Greek text. The textual evidence, which involves four 
variant readings, is as follows: 

(1) 'grace' (xapiros) is read by p 74 X A B D, more than 
twenty minuscule manuscripts, the Vulgate, Sahidic, 
Bohairic, Syriac p , Armenian, and Ethiopic (the last reads 
\dpiros 9eov). 

(2) 'faith' (mcTTeajj) is read by H a P a , many minuscule manu- 
scripts, Syriac h , and Chrysostom. 

(3) 'grace and faith' (xdpiTos ko.1 mareo)?) is read by E\ 

(4) 'faith and grace of the Spirit' (morecos ko.1 %dpiTos nvev- 
jxaros) is read by W. 

Of these four variant readings it is obvious that either the 
first two are independent abridgements of the longer readings, 
or the third and fourth readings have arisen from combining the 
elements of the first two. Considerations of both external evi- 
dence and internal probability unite to demonstrate that read- 
ings (3) and (4) are secondary, being alternative conflations of 
the other two. Reading (3) is supported by the uncial manu- 
script E a , which dates from the sixth century and is one of the 
earliest representatives of the Koine or Byzantine type of text 

222 The Practice of New Testament Textual Criticism 

in Acts. Reading (4) is supported by the uncial manuscript W, 
which dates from the eighth or ninth century and has a mixed 
type of text in Acts. Transcriptional considerations lead one to 
conclude that both (3) and (4) presuppose the priority of the 
other two readings, for it is easier to believe that a scribe, know- 
ing the existence of readings (1) and (2), decided to join them, 
lest the copy which he was writing lose one or the other, than to 
believe that two scribes independently took offence at the longer 
reading and that each chose to perpetuate half of it in his copy. 
Thus external evidence, which is meagre in extent and rela- 
tively late in date, and transcriptional probabilities unite against 
the originality of readings (3) and (4). 

Variant reading (2) is supported by two uncial manuscripts, 
H a of the ninth century and P a of the tenth century, both 
representative of the Koine or Byzantine type of text. The 
majority of the minuscule manuscripts join these two uncial 
witnesses. The earliest witness to reading (2) is Chrysostom, who 
died a.d. 407. 

Variant reading (1) is supported by a wide variety of wit- 
nesses, including representatives of the major pre-Koine types 
of text. Codices Sinaiticus and Vaticanus, both of the fourth 
century, are the earliest and best uncial representatives in Acts 
of the Alexandrian type of text. Codex Bezae, of the fifth or 
sixth century, is the chief Greek representative of the Western 
group of witnesses. Codex Alexandrinus, of the fifth century, 
and p 74 , dating from about the seventh century, have a mixed 
type of text. The evidence of the early versions, including the 
Latin, Syriac, Coptic, and Armenian, reflect the wide geo- 
graphical area over which the reading was accepted. The 
external evidence in support of reading (1) is, therefore, far 
superior in point of age and diversity of text-type to that 
supporting reading (2). 

Internal probabilities likewise favour reading (1). If the 
account originally stated that Stephen was 'full of faith', there 
is no discernible reason why a scribe should alter it to 'full of 
grace'. On the other hand, in view of the statement made three 
verses earlier that Stephen was a man 'full of faith and the Holy 
Spirit' (vs. 5), it is easy to understand that in transcribing the 
later statement in verse 8 copyists would be likely, either con- 
sciously or unconsciously, to substitute moreens, which they 

The Practice of New Testament Textual Criticism 223 

recalled from the earlier passage, for the correct reading xa-piros. 
The presence of Trvev^aros in reading (4) is to be explained in 
the same way. 

Thus the converging of several strands of evidence, both 
external and internal, leads one to the firm conclusion that the 
author of Acts vi. 8 wrote nX-qp-qs xaptro? xal Swdfieajs. 

Not a few New Testament manuscripts incorporate here and 
there interesting details, some of which may be historically cor- 
rect. The story of the woman taken in adultery, for example, has 
all the earmarks of historical veracity ; no ascetically minded 
monk would have invented a narrative which closes with what 
seems to be only a mild rebuke on Jesus' part: 'Neither do I con- 
demn you; go, and do not sin again.' At the same time the peri- 
cope, which is usually printed as John vii. 53-viii. 1 1 , must be 
judged to be an intrusion into the Fourth Gospel. 

The account is lacking in the best Greek manuscripts: it is 
absent from p 66 p 75 X B L N T W X A & W 33 157 565 892 1241 
fam. 1424, &c. Codices A and C are defective at this point, but 
it is highly probable that neither contained the section, for there 
would not have been space enough on the missing leaves to in- 
clude it along with the rest of the text. The Old Syriac (Syr 3 - c ) 
and the Arabic form of Tatian's Diatessaron betray no know- 
ledge of the passage, nor is it contained in the best manuscripts of 
the Peshitta. Likewise the old Coptic Churches did not include it 
in their Bible, for the Sahidic, the sub-Achmimic, and the older 
Bohairic manuscripts lack it. Some Armenian manuscripts as 
well as the Old Georgian version omit it. In the West the passage 
is absent from the Gothic version and from several Old Latin 
manuscripts (afl*q). 

Even more significant is the fact that no Greek Church 
Father for a thousand years after Christ refers to the pericope, 
including even those who, like Origen, Chrysostom, and Nonnus 
(in his metrical paraphrase) , dealt with the entire Gospel verse 
by verse. Euthymius Zigabenus, who lived in the first part of 
the twelfth century, is the first Greek writer to comment on the 
passage, and even he declares that the accurate copies of the 
Gospel do not contain it. 

When one adds to this impressive and diversified list of ex- 
ternal evidence the consideration that the style and vocabulary 

224 The Practice of New Testament Textual Criticism 

of the pericope differ markedly from the rest of the Fourth Gos- 
pel, and that it interrupts the sequence of vii. 52 and viii. 12 f., 
the case against its being of Johannine authorship appears to 
be conclusive. 

The earliest Greek manuscript known to contain the passage 
is codex Bezae, of the fifth or sixth century, which is joined by 
several Old Latin manuscripts (bceff 2 j). The pericope is 
obviously a piece of floating tradition which circulated in 
certain parts of the Western Church. It was subsequently in- 
serted into various manuscripts at various places. Most scribes 
thought that it would interrupt John's narrative least if it were 
inserted after vii. 52 (D E F G H K M S U FA 77 28 579 700 
1579, &C-). Others placed it after vii. 36 (MS. 225) or after 
xxi. 24 (fam. 1 1076 1570 1582). The revision of the Old Georgian 
version, made in the eleventh century by George the Athonite, 
contains the passage after vii. 44. The scribe of an ancestor of 
fam. 13 inserted it in another Gospel altogether, after Luke 
xxi. 38. Significantly enough, in many of the manuscripts which 
contain the passage it is marked with an obelus (as, for example, 
in S) or an asterisk (as, for example, in E M A), indicating that, 
though the scribes of these manuscripts included the account, 
they were aware that it lacked satisfactory credentials. 

A few of the manuscripts which report the incident also in- 
clude an interesting expansion at the close of viii. 8. More than 
one reader of the statement that Jesus 'bent down and wrote 
with his finger on the ground' must have wondered what it was 
that the Lord wrote. An unknown copyist satisfied this natural 
curiosity by adding the words, 'the sins of each of them'. 1 

The best disposition to make of the pericope as a whole is 
doubtless to print it at the close of the Fourth Gospel, with a 
footnote advising the reader that the text of the pericope has 
no fixed place in the ancient witnesses. 2 

An interesting example of several different attempts to clarify 
a comment made by the Fourth Evangelist, which was felt to 
be open to misinterpretation, is found in John vii. 37-39: 

1 See David C. Voss, 'The Sins of Each One of Them', Anglican Theological 
Review, xv (1933), pp. 321-3. 

2 For a discussion of the origin of the pericope and the problem of why it was 
so late in being incorporated into the canonical text, see Harald Riesenfeld, 'Peri- 
kopen de adultera i den fornkyrkliga traditionen', Svensk exegetisk drsbok, xvii (1953), 
pp. 106-18. See below, p. 273. 

The Practice of New Testament Textual Criticism 225 

On the last day of the feast, the great day, Jesus stood up and 
proclaimed, 'If any one thirst, let him come to me and drink. He who 
believes in me, as the scripture has said, "Out of his heart shall flow 
rivers of living water".' Now this he said about the Spirit, which those 
who believed in him were to receive ; for as yet the Spirit had not been 
given, because Jesus was not yet glorified. 

In the final sentence the clause 'for as yet the Spirit had not 
been given' (owtu yap rjv Trvevfxa SeSo/xeVov) appears in seven dif- 
ferent forms : 

(1) irvevixa p<* c p^s x K T IJ V 1079 1546 Cop bo Arm. 

(2) TTvevfjLa dyiov p 66 * L W X T A A 28 33 565 700. 

(3) nvev^ux ayiov c'tt' aureus D^Goth. 

(4) TTvevna 8e8o/j,evov a b c e ff 1 g I Vulg Syr E * c> p Eusebius. 

(5) irvevfMa ayiov 8e8ofj.evov B 053 1230 e q Syr pal * h . 

(6) 'for they had not yet received [the] Spirit' Cop sah - sub - ach - 

(7) 'for the Holy Spirit had not yet come' Eth. 

A little reflection will make it obvious that the reading which 
explains the rise of all the others is ( 1 ) nvevixa. Many scribes were 
doubtless perplexed by the bare and ambiguous statement 'for 
as yet the Spirit was not, because Jesus was not yet glorified'. Lest 
this be taken to affirm that the Spirit was not yet in existence 
prior to Jesus' glorification, modifications were introduced to re- 
lieve the difficulty. Several Western witnesses (D/Goth) read (3), 
'for the Holy Spirit was not yet in them'. Other witnesses add 
the verb 'given' (as in readings 4 and 5), or 'received' (read- 
ing 6), or 'come' (reading 7). 

The introduction of the adjective ayiov (readings 2, 3, and 5) 
is a most natural kind of addition that many scribes would make 
independently of one another. (The correction found in p 66 , de- 
leting ayiov, is in keeping with the observed vigilance of this 
scribe in correcting his own inadvertent errors.) It is noteworthy 
that in this case codex Vaticanus is doubly in error (5), having 
added both ayiov and a predicate verb. 

The evidence for (2) can be joined to that of (1) in respect 
of resisting the temptation to add a predicate verb. There is thus 
a very widespread and diversified constellation of witnesses in 
support of the more difficult and shorter reading. It can scarcely 
be doubted, therefore, that the original text had simply ovmco yap 
7]v TTvevjxa. 1 

1 This is the text lying behind the King James version, the Revised Version of 

226 The Practice of New Testament Textual Criticism 

How did Mark end his Gospel? Unfortunately, we do not 
know ; the most that can be said is that four different endings are 
current among the manuscripts, but that probably none of them 
represents what Mark originally intended to stand as the close 
of his Gospel. These four endings may be called the short ending, 
the intermediate ending, the long ending, and the long ending 
expanded. The evidence for each of them is as follows : 

( i ) The last twelve verses of Mark (xvi. 9-20) are lacking in 
the two earliest parchment codices, B and X, in the Old Latin 
manuscript A:, the Sinaitic Syriac, many manuscripts of the Old 
Armenian version, the Adysh and Opiza manuscripts of the Old 
Georgian version, and a number of manuscripts of the Ethiopic 
version. Clement of Alexandria, Origen, and Ammonius show no 
knowledge of the existence of these verses; other Church Fathers 
state that the section is absent from Greek copies of Mark known 
to them (e.g. Jerome, Epist. cxx. 3, ad Hedibiam, 'Almost all the 
Greek copies do not have this concluding portion'). The origi- 
nal form of the Eusebian sections makes no provision for number- 
ing sections after xvi. 8. Not a few manuscripts which contain 
the passage have scholia stating that older Greek copies lack it 
(so, for example, MSS. 1, 20, 22, &c), and in other witnesses 
the passage is marked with asterisks or obeli, the conventional 
sigla used by scribes to indicate a spurious addition to a literary 

(2) The intermediate ending ('But they reported briefly to 
Peter and those with him all that they had been told. And after 
this Jesus himself sent out by means of them, from east to west, 
the sacred and imperishable proclamation of eternal salvation') 
is present in several uncial manuscripts of the seventh, eighth, 
and ninth centuries (L fogg 01 12), as well as in a few minus- 
cule manuscripts (274 mg [see Plate XI] 57g), and several ancient 
versions (k Syr hm « Coptic pt Eth codd ). 1 

(3) The long ending, so familiar through the King James 
version and other translations of the Textus Receptus, is present 

1881, and the American Standard Version of 1901, all of which use italics to show 
what the translators added for the sake of English readers: '. . . was not yet given'. 
Since neither the Revised Standard Version nor the New English Bible employs 
italics in this way, the inclusion of the verb 'given' in these two versions is to be 
accounted for either as the result of licence in translation or the choice of what 
appears to be a secondary variant reading as the basic text. 

1 For detailed evidence of the Coptic versions, see P. E. Kahle, Journal of 
Theological Studies, N.s., ii (1951), pp. 49-57- 

The Practice of New Testament Textual Criticism 227 

in the vast number of witnesses (including several which also 
contain the intermediate ending), namely A C D L W @, most 
of the later uncials, the great majority of the minuscules, most 
of the Old Latin witnesses, the Vulgate, Syr c> p , and Coptic pt . 
It is probable that Justin Martyr at the middle of the second 
century knew this ending ; in any case Tatian, his disciple, in- 
cluded it in his Diatessaron. 

(4) The long ending in an expanded form existed, so Jerome 
tells us, in Greek copies current in his day, and since the dis- 
covery of W earlier this century we now have the Greek text of 
this expansion (for a translation of the addition after vs. 14, see 
p. 57 above). 

None of these four endings commends itself as original. The 
obvious and pervasive apocryphal flavour of the expansion in (4) , 
as well as the extremely limited basis of evidence supporting 
it, condemns it as a totally secondary accretion. 

The long ending (3), though present in a variety of witnesses, 
some of them ancient, must also be judged by internal evidence 
to be secondary. For example, the presence of seventeen non- 
Marcan words or words used in a non-Marcan sense ; the lack of 
a smooth juncture between verses 8 and 9 (the subject in vs. 8 
is the women, whereas Jesus is the presumed subject in vs. 9) ; 
and the way in which Mary is identified in verse 9 even though 
she has been mentioned previously (vs. 1) — all these features 
indicate that the section was added by someone who knew a 
form of Mark which ended abruptly with verse 8 and who 
wished to provide a more appropriate conclusion. An Armenian 
manuscript of the Gospels, copied a.d. 989 (see Plate XlVi), 
contains a brief rubric of two words in the space at the end of 
the last line of verse 8 and before the last twelve verses, namely 
Ariston eritsou ('of the Presbyter Ariston'). Many have interpreted 
this as a reference to Aristion, a contemporary of Papias in 
the early second century and traditionally a disciple of John 
the Apostle. But the probability that an Armenian rubricator 
would have access to historically valuable tradition on this point 
is almost nil, especially if, as has been argued, the rubric was 
added in the thirteenth or fourteenth century. 1 

The internal evidence of the so-called intermediate ending (2) 
is decidedly against its being genuine. Besides containing a high 

1 So Joseph Schafers in Biblische £eitschrifl, xiii (1915), pp. 24-25. 

228 The Practice of New Testament Textual Criticism 

percentage of non-Marcan words, its rhetorical tone differs 
totally from the simple style of Mark's Gospel. The mouth- 
filling phrase at the close ('the sacred and imperishable mes- 
sage of eternal salvation') betrays the hand of a later Greek 

Thus we are left with the short ending, witnessed by the 
earliest Greek, versional, and patristic evidence. Both external 
and internal considerations lead one to conclude that the 
original text of the Second Gospel, as known today, closes at 
xvi. 8. But did Mark intend to conclude his Gospel with the 
melancholy statement that the women were afraid (e'<£o/3owro 
yap)? Despite the arguments which several modern scholars 
have urged in support of such a view, 1 the present writer can- 
not believe that the note of fear would have been regarded as 
an appropriate conclusion to an account of the Evangel, or 
Good News. Furthermore, from a stylistic point of view, to 
terminate a Greek sentence with the word yap is most unusual 
and exceedingly rare — only a relatively few examples have been 
found throughout all the vast range of Greek literary works, and 
no instance has been found where yap stands at the end of a book. 
Moreover, it is possible that in verse 8 Mark uses the verb 
ifofiovvTo to mean 'they were afraid of (as he does in four of the 
other occurrences of this verb in his Gospel). In that case 
obviously something is needed to finish the sentence. 

It appears, therefore, that ifofiovvTo yap of Mark xvi. 8 does 
not represent what Mark intended to stand at the end of his 
Gospel. Whether he was interrupted while writing and sub- 
sequently prevented (perhaps by death) from finishing his 
literary work, or whether the last leaf of the original copy was 
accidentally lost before other copies had been made, we do not 
know. All that is known is that more than one person in the 
early Church sensed that the Gospel is a torso and tried in 
various ways to provide a more or less appropriate conclusion. 2 

1 e.g. J. M. Creed, Journal of Theological Studies, xxxi (1932), pp. 175-80; Ernst 
Lohmeyer, Das Evangelium des Markus (Gottingen, 1937), pp. 356-60; N. B. Stone- 
house, The Witness of Matthew and Mark to Christ (Philadelphia, 1944), pp. 86-1 18; 
A. M. Farrer, The Glass of Vision (London, 1948), pp. 136-46; R. H. Lightfoot, 
The Gospel Message of St. Mark (Oxford, 1950), pp. 80-97, 106-16. 

1 Almost all textual studies and critical commentaries on the Gospel according 
to Mark agree that the last twelve verses cannot be regarded as Marcan; typical is 
the monograph by Clarence R. Williams, The Appendices to the Gospel according to 

The Practice of New Testament Textual Criticism 229 

It should not be overlooked that the text-critical analysis of 
the endings of Mark's Gospel has an important bearing on the 
historical and literary source criticism of the Gospels. Since 
Mark was not responsible for the composition of the last twelve 
verses of the generally current form of his Gospel, and since they 
undoubtedly had been attached to the Gospel before the Church 
recognized the fourfold Gospels as canonical, it follows that the 
New Testament contains not four but five evangelic accounts of 
events subsequent to the Resurrection of Christ. 

In contrast with the length of the preceding textual variant, 
embracing twelve verses, one or two variant readings will now 
be examined which involve the presence or absence of a single 

The Greek text which lies behind the traditional words of 
the angelic chorus at the birth of Jesus ('Glory to God in the 
highest, and on earth peace, good will toward men', Luke ii. 14, 
King James version) differs by only the letter sigma from the 
Greek text which lies behind the same verse in the Revised 
Standard Version ('Glory to God in the highest, and on earth 
peace among men with whom he is well pleased!'). In the 
former case the King James translators followed the Textus 
Receptus, which reads evSoKia, supported by EFGHLSU 
V & S Q, many minuscules, the Syriac, Bohairic, and Geor- 
gian versions, Tatian, and Eusebius. In the latter case the 
Revisers followed a Greek text which reads the genitive case, 
evSoxtas, supported by B* X* A D W 28, the Old Latin, 
Jerome's Vulgate, the Gothic, the Sahidic, Irenaeus, Cyril of 
Jerusalem, and the Latin Fathers. 

Mark: a Study in Textual Transmission (= Transactions of the Connecticut Academy of Arts 
and Sciences, vol. xviii; New Haven, 191 5). 

In the last century two major attempts were made to defend the genuineness of 
the long ending ; namely John W. Burgon's The Last Twelve Verses of the Gospel accord- 
ingtoSt. Mark Vindicated . . . (London, 1871; reprinted 1959), andj. P. P. Martin's 
Introduction a la critique textuelle du Nouveau Testament: Partie pratique, vol. ii (Paris, 
1884). The most recent effort to rehabilitate this position was made by M. van der 
Valk in his 'Observations on Mark 16, 9-20 in Relation to St. Mark's Gospel', 
published in Humanitas, N.s., vi-vii (Instituto de estudos classicos, Faculdade de 
letras da Universidade de Coimbra, 1958), pp. 52-95. This essay is a singular 
exhibition of how not to practise textual criticism! Omitting entirely all considera- 
tion of the external evidence of manuscripts, van der Valk concentrates on more 
or less irrelevant and speculative considerations in order to arrive at what appears 
to be a predetermined conclusion. See below, p. 273. 

230 The Practice of New Testament Textual Criticism 

It will be observed that the earliest Greek manuscripts of both 
the Alexandrian and Western groups are joined by significant 
versional and patristic evidence in support of evSoKias, proving 
that this reading was widely disseminated in the West and was 
known in Upper Egypt and in Palestine as well. On the other 
hand, the nominative form of the word was also widely read, 
being current in Syria, in Egypt around the delta, at Caesarea 
(0, Eusebius) , and dating from as early as the second century 
(Tatian) . On the whole, the external evidence appears to favour 
the genitive case, for the combination of B X W with D and the 
Old Latin outweighs the external support for the nominative 
case. Internal considerations confirm this judgement, for in the 
context the genitive case is the more difficult to construe and 
therefore would be more likely to be altered to the nominative 
than vice versa. Furthermore, consideration of intrinsic suit- 
ability corroborates the other evidence, for the expression 'men 
of [his, i.e. God's] good will [or favour]' is a perfectly good 
Semitic construction, appearing several times in the Hebrew 
hymns discovered at Qumran, 1 and thus is entirely congruous 
with the Semitic cast of the first two chapters of Luke. 

Luke ii. 14, therefore, involves a twofold strophe ('glory . . . 
peace'), and not a threefold strophe ('glory . . . peace . . . good 
will', all in the nominative). The sense is that the birth of the 
Messiah, the Lord (vs. 11), is the occasion for the ascription of 
glory to God in highest heaven and the enjoyment of peace on 
earth among men of God's good will, i.e. those persons on whom 
his favour rests, chosen to be the recipients of the gift of his 

Another interesting variant reading which arises from the 
presence or absence of a single Greek letter is found in 1 Thess. 
ii. 7. Here the reading is either 'we were gentle (yjmoi) among 
you, like a nurse taking care of her children', or 'we were babes 
(vrpnoi) among you, like a nurse taking care of her children'. 
The word 1777101 is supported by N c A C b D p(c) K? L? P p 33 
Syr p ' h ,Sahidic, Armenian, Clement 2/2, Origen3/4, Chrysostom, 

1 See Claus-Hunno Hunzinger, 'Neues Licht auf Lc. 2. 14', ^eitschrift fur die 
neutestamentliche Wissenschaft, xliv (1952-3), pp. 85-90; id., 'Ein weiterer Beleg zu 
Lc. 2. 14 avBpumoi. tv&OKtas', ibid, xlix (1958), pp. 129-31, and cf. Ernest Vogt, 
' "Peace among Men of God's Good Pleasure" Lk. 2. 14', in The Scrolls and the New 
Testament, ed. by K. Stendahl (New York, 1957), pp. 1 14-17. See below, p. 273. 

The Practice of Mew Testament Textual Criticism 231 

Theodore of Mopsuestia ; the word vqirwi is supported by p 6s 
X* B C* D p * Fp Gp I, Old Latin, Vulgate, Bohairic, Ethiopic, 
Origen ,at , Ephraem, Cyril, Jerome, Augustine, and Ambrosi- 

It is easy to see how each of these variant readings may have 
arisen, for the word which precedes the variant is iyeirqOTj/xev. 
When the text was copied by dictation, the pronunciation 
of iyevrjOrjiiev 1777-101 would be almost indistinguishable from 
iyevTjOrjiJLev vtjttioi, and vice versa. Likewise when the words 
stand in uncial script, ErENH0HMENNHIIIOI is very like 
ErENHQHMENHIIIOI, and vice versa. Furthermore, it must 
be remembered that N at the end of a line was frequently in- 
dicated merely by a stroke above the preceding letter, thus: 

It is less easy to decide in which direction the change went. 
On the one hand, the weight and diversity of external evidence 
are clearly in favour of vrjmoi, which is supported by the earliest 
form of the Alexandrian text (p 65 [third century], X*, and B), 
the Western text (D* and Old Latin), as well as a wide variety 
of Versions and Fathers. Such a constellation of witnesses led 
Lachmann, Westcott-Hort, Zimmer, 1 and Bover to print vrjmoi. 
in their text, a choice favoured also by such commentators as 
Lightfoot, Findlay (doubtfully) , Wohlenberg, Frame, and Milli- 
gan (doubtfully). 

On the other hand, Paul's violent transition in the same 
sentence from a reference to himself as a babe to the thought of 
his serving as a mother-nurse has seemed to most editors and 
commentators to be little short of absurdity, and therefore Tre- 
gelles, Tischendorf, Alford, Weiss, von Soden, Merk, and Vogels 
follow the Textus Receptus and print -qmoi, a reading which 
is supported by Bornemann, von Dobschutz, Moffatt, Dibelius, 
Lemonnyer, Voste, Neil, and Rigaux, and is adopted by the 
translators of the Revised Standard Version and the New 
English Bible. 

Those who are impressed by the external evidence supporting 
vrjmoi attempt to alleviate the difficulty in sense that this word 
introduces by pointing to the earlier context (vss. 3-6), where 

1 Besides his monograph on Der Text der Thessalonicherbrie/e (Quedlinburg, 1 893) , 
Friedrich Zimmer supported his view in his contribution to Bernhard Weiss's 
Festschrift, Theologische Studien (Gottingen, 1897), pp. 248-73, especially pp. 264-9. 

232 The Practice of New Testament Textual Criticism 

Paul speaks not of his gentleness but of his unselfish love. He 
defends himself against the charge, not of being harsh, but 
of using flattery to gain an unfair advantage. He suggests his 
lack of guile and childlike innocence, so it is argued, by calling 
himself a 'babe'. The word for 'babe' by association brings with 
it the idea of a 'nurse', and, with his characteristic rapidity of 
thought, Paul inverts the metaphor and now refers to himself 
as a mother-nurse. Such an inversion of metaphor — a Chris- 
tian teacher being first compared to the child and then to the 
mother — is quite in the apostle's manner; e.g., in Gal. iv. 19 
the sudden shift in metaphor is even more startling ('My little 
children, with whom I am again in travail until Christ be 
formed in you!'). It is also significant that twelve of the sixteen 
cases of 1^7710? in the New Testament are found in Paul (including 
Eph. iv. 14), whereas 177710? appears elsewhere in the Greek Bible 
only in 2 Tim. ii. 24. 

Not all of these arguments in favour of vr/moi, however, are 
as significant as they may seem to be at first sight. Whatever may 
be said of the more remote context, in the immediate context 
'gentle' is certainly the appropriate antithesis to what Paul has 
just disclaimed in verse 6b, namely the assertion of his apostolic 
dignity or authority, while in connexion with the following 
clauses it is immeasurably preferable to 'babes'. Furthermore, 
though it is true that Paul uses ^777710? frequently, it must not be 
overlooked that he never refers to himself as a babe, but always 
applies it to his converts. Moreover, it is certainly not without 
importance that at 2 Tim. ii. 24 more than one scribe suc- 
cumbed to the temptation to substitute the more familiar word, 
vfyrios, for the true text 177710? (so the Greek text of the four 
manuscripts D?* E p * F? G? and the Ethiopic). 1 

By way of striking a balance in the arguments pro et contra, 
it appears that here internal considerations should be allowed 
to take precedence over external evidence. Since 1^777101, which 
is by far the more common word in the New Testament, could 
so easily have been introduced by carrying over the final v of the 
preceding word, and since the violence done to the sense' when 
1^777101 is read appears to be intolerable, one is entitled to apply 
Daniel Mace's crisp dictum that no manuscript is so old as 

1 The opposite error of writing rjmos for vrjmos also occurs sporadically in 
individual manuscripts, namely in A at Eph. iv. 14 and in 33 at Heb. v. 33. 

The Practice of New Testament Textual Criticism 233 
common sense, and to make a tentative decision in favour of 


A knowledge of Hebrew and especially of Aramaic will occa- 
sionally throw light upon a variant reading in the Gospels. For 
example, the words of Jesus in Mark xiv. 25, 'Truly I say to you, 
I shall not drink again of the fruit of the vine until that day when 
I drink it new in the kingdom of God', are transmitted in three 
different forms. The text which lies behind the English versions 
is (1) ovkcti oil /jL.rj ttloj, supported by AB A, most minuscules, 
bff z i I q r z , Vulg, Syr s,p ' h , Sab., Geo, (Arm). There are, how- 
ever, two other readings: (2) ov pji) ttloj, supported by X C (D) 
L W 471 892 1342 c k, Vulg (1 MS.), Boh, Eth; and (3) ov fir) 
Trpoo-dG) {tt€iv), supported by D © 565 af Arm. (@ reads 

OVKiTl OV /XT] TTpoa9u>ll€V TTl€lV.) 

The third reading, found in Western and Caesarean witnesses, 
involves a Semitic idiom, meaning literally, 'I shall not add to 
drink . . .'. This idiom appears frequently in the Septuagint, 
which has more than 100 examples of -rrpoo-TiOevai (rrpoo-Tideo-dat) 
for the Hebraic FpDin (with an infinitive), meaning tto\lv. It 
appears that the Eucharistic words of Jesus, which were un- 
doubtedly spoken in Aramaic or Hebrew to the apostles, have 
been preserved in literalistic fashion in the third variant 
reading, whereas the other two readings provide alternative 
interpretations of the meaning, expressed in more idiomatic 
Greek. (The second reading, in fact, can be called a misinter- 
pretation, for it omits the idea expressed by 'again'.) Whether 
the idiom which is preserved in the third variant reading is to be 
considered original to Mark's Gospel (as Wellhausen 1 suggested), 
or whether it is a secondary Biblicism in imitation of the Sep- 
tuagint (as Schiirmann 2 argues), or whether it discloses influence 
from extra-canonical reports of Jesus' Eucharistic words (as 
Jeremias, 3 following Black, 4 prefers), it is in any case a most 

1 Julius Wellhausen, Einleitungin die drei ersten Evangelien, 2te Aufl. (Berlin, 191 1), 


2 Heinz Schiirmann, Der Paschamahlbericht Lk. 22, (7-14) 15-18 (Munster/W., 
'953). P- 35. Anm. 154. 

3 Joachim Jeremias, Die Abendsmahlsworte Jesu, 3te Aufl. (Gottingen, i960), 
pp. 174 f. ; Eng. trans., The Eucharistic Words of Jesus (London, 1966), p. 183. 

4 Matthew Black, An Aramaic Approach to the Gospels and Acts, 2nd ed. (Oxford, 
•954). PP- 2!4 f -; 3 rd ed. (1967), pp. 279 f. 

234 The Practice of New Testament Textual Criticism 

noteworthy variant reading which is eminently suitable in the 

Many scholars 1 have given attention to the textual problem 
found in Acts xx. 28, the report of Paul's farewell address to the 
Ephesian elders : 'Take heed to yourselves and to all the flock, 
in which the Holy Spirit has made you guardians, to feed the 
church of . . . which he obtained with his own blood.' Should 
the ellipsis be filled with the words 'the Lord' or 'God'? 

The external evidence is as follows : 

(1) deov X B 218 257 383 459 614 917 1 175 1522 161 1 1758 

2138 2298, al, Vulg, Syr h , Basil, Ambrose, Epiphanius, 
Cyril of Alexandria. 

(2) icvpiov p™ A C* D E a W 33 181 209 307 337 429 431 436 

610 623 1739 1891, gig, Sah, Boh, Syr hmg , Arm, 
Irenaeus lat , Didymus, Lucifer. 

(3) Kvpiov Kal deov C 3 H a L a P a , more than 100 minuscules, 

Old Slavonic, Theophylact. 

(4) deov Kal Kvpiov 47. 

(5) kv p' iov eeoC 3 95**- 

(6) Xpiarov Syr codd , Apost. Const., Athanasius codd . 

(7) 'Irjaov Xpiarov tn. 

Of these seven readings, obviously only the first two merit any 
attention, for (6) and (7) are insufficiently supported, and the 
rest are conflations of Kvpiov and deov in various combinations 
preserved in Byzantine texts. 

On the basis of external evidence it is difficult to decide which 
of the first two readings is original. Palaeographically the 
difference concerns only a single letter: KY and 0Y. Each is 
supported by early and diversified witnesses. Perhaps the most 
that can be said is that deov is the Alexandrian reading and 
Kvpiov is supported by typical Western documents. One must 
rely chiefly on considerations of internal probabilities in reach- 
ing a decision. 

1 See, for example, Ezra Abbot, 'On the Reading "Church of God", Acts xx. 28', 
Bibliolheca Sacra, xxxiii (1876), pp. 313-52 (reprinted in the volume The Authorship 
of the Fourth Gospel and Other Critical Essays [Boston, 1888], pp. 294-331), Charles F. 
De Vine, 'The "Blood of God" in Acts 20:28', Catholic Biblical Quarterly, ix (1947), 
pp. 381-408, as well as critical commentaries on the passage and the discussion by 
Hort in The New Testament in the Original Greek, [iij Appendix, pp. 98-100. 

The Practice of New Testament Textual Criticism 235 

The phrase ewwAijata Kvplov occurs seven times in the Septua- 
gint but nowhere else in the New Testament. On the other hand, 
€KK\r]oLa tov deov appears with moderate frequency (eleven 
times) in the Epistles traditionally ascribed to Paul and no- 
where else in the New Testament. (The expression at e/c/cA^atai 
■naaai tov Xpiarov occurs once in Rom. xvi. 16.) It is possible, 
therefore, that a scribe, finding deov in his exemplar, was in- 
fluenced by Old Testament usage and altered it to Kvpiov. On 
the other hand, it is also possible that a scribe, influenced by 
Pauline usage, changed xvpiov of his exemplar to deov. 

Two considerations which, in the opinion of the present 
writer, tip the scales in favour of the originality of the reading 
deov are the following : 

(a) It is undeniable that deov is the more difficult reading. 
The following clause refers to the church 'which he obtained 81a 
tov aifiaros tov 181W. If this is taken in its usual sense ('with his 
own blood'), the scribe might well raise the question, Does God 
have blood?, and thus be led to change deov to xvpiov. On the 
other hand, if Kvpiov were the original reading, there is nothing 
unusual in the context which would cause a scribe to introduce 
the more difficult deov. 

(b) The other consideration asks which reading is more likely 
to have been altered during the Arian controversy that raged 
over the Person of Christ. In this connexion Alford's reasoning 
still seems to be cogent. He writes: 

If the passage is of such a nature that, whichever reading is 
adopted, the orthodox meaning is legitimate, but the adoption of 
the stronger orthodox reading is absolutely incompatible with the 
heretical meaning, — then it is probable that such stronger orthodox 
reading was the original. For while the heretics would be certain 
[the present writer would prefer to say 'tempted'] to annul the 
expression offensive to them and substitute the weaker one, the 
orthodox, on the above hypothesis, would have originally no motive 
for alteration. 1 

In paragraph (a) above it was implied that 8«x tov at/xaTos 
tov ISiov could have a meaning other than its usual one ('with 
his own blood') ; this other meaning, which may have been in 
the mind of the writer of Acts, is 'with the blood of his Own'. 

1 Henry Alford, The Greek Testament, new ed., i (Boston, 1881), p. 83 of the pro- 
legomena, n. 1. 

236 The Practice of New Testament Textual Criticism 

(It is not necessary to suppose, with Hort, that vlov may have 
dropped out after tov ISlov, though palaeographically such 
an omission would have been easy.) This absolute use of the 
singular number of 1810s, which is otherwise unknown in the 
New Testament, is found occasionally in the Greek papyri as 
a term of endearment referring to near relatives. 1 It is pos- 
sible, therefore, that 'his Own' (o i'Sio?) was a title which early 
Christians gave to Jesus, comparable with 'the Beloved' (o 
ayanrjTos) ; compare Rom. viii. 32, where Paul refers to God 
'who did not spare tov 181ov vlov' in a context that clearly 
alludes to Gen. xxii. 1 6, where the Septuagint has tov aya-n-qTov 


In conclusion, whatever may be thought of the slight possi- 
bility that tov ISlov is used here as the equivalent of tov 181ov 
vlov, it appears that the reading 9eov was more likely to have 
been altered to Kvpiov than vice versa, and it is therefore to be 
regarded as the original. 2 

Of all the letters attributed to Paul, his letter to the Colossians 
contains proportionately the greatest number of textual prob- 
lems. The close of Col. ii. 2 presents what is, at first, a bewilder- 
ing variety of readings; the manuscripts present fifteen different 
conclusions of the phrase els iiriyvcjaiv tov fj,voTr]plov . . . ('to 
the knowledge of the mystery of. . .') : 

A (1) tov deov XpioTov p« 6 B and Hilary of Poitiers. 

(2) tov deov Hp Pp 69 424** 436 462 191 2, Sahidic Bea "y MS - 

(3) TOV XpLOTOV I462. 

(4) tov deov 6 eoTuv Xpioros D p * d e, Augustine. 

(5) tov deov o eonv irepl XpioTOV Ethiopic. 

(6) tov deov tov ev XpiaTOj 33, Clement. 

(7) tov deov tov ev Xpiarco 'Irjaov Armenian. 

(8) tov deov Kal XpioTov Cyril of Alexandria. 
B (9) tov deov -naTpos XpiaTov X* 2 16 44°- 

1 James Hope Moulton, A Grammar of New Testament Greek, i, Prolegomena, 3rd ed. 
(Edinburgh, 1908), p. 90. 

1 The reading deov is preferred by Westcott-Hort, Weiss, Vogels, Merk, Bover, 
Harnack (Beitrage, iv, p. 330), Jackson and Lake (in Beginnings of Christianity), 
G. H. C. Macgregor (in Interpreter' s Bible) , F. F. Bruce (Acts), C. S. C. Williams (in 
Harper's [= Black's] Mew Testament Commentaries) , and De Vine (op. cit.). 

The reading Kvpiov is preferred by Tischendorf, Abbot, von Soden, Ropes, the 
Revised Standard Version, and the New English Bible. 

The Practice of New Testament Textual Criticism 237 

(10) tov deov TraTpos tou XpiCTTov A C 4, Sahidic, Bohairic. 

( 1 1 ) tov deov Kal TraTpos tov XpioTov X c , Syr h . 

(12) tov deov TraTpos Kal tov XpioTov 441 1908, Syr p , 


(13) tov deov TraTpos Kal tov XpioTov 'Irjcrov Vulgate, 

Bohairic codd , Pelagius. 

(14) tov deov TraTpos Kal tov Kvpiov rjfjxbv XpioTov 'Irjoov 

i Vulgate MS. 

(15) tov deov Kal TraTpos Kal tov XpiaTov D p ' c ' E p K p L p , 

most minuscules, Theodoret, &c. 

Of all these variant readings the one which has been placed 
first is to be preferred on the basis of both external and internal 
considerations. Externally, it is supported by the earliest and 
best Greek manuscripts ; internally, the difficulty of interpreting 
the meaning of the expression tov fjLvoTrjplov tov deov XpioTov 
has led to the multiplication of scribal attempts to clarify the 
sense. An obviously popular expedient was the insertion of the 
word TraTpos; this addition appears in seven of the variant read- 
ings (those grouped under B). The insertion of the article 
before XpiaTov (readings 10-15) ^ s plainly in the interest of 
making the expression parallel with tov deov. The reading 
placed last in the list (it lies behind the rendering of the King 
James version, 'the mystery of God, and of the Father, and of 
Christ'), though supported by the largest number of witnesses, 
is also the weakest, for it is a conflation of the two types of 
amelioration represented in (11) and (12). 

If reading (9) were original, then the rise of all eight readings 
grouped under A is inexplicable, for why should TraTpos have 
fallen out? On the contrary, TraTpos was inserted in order to 
clarify the syntactical relation between 6eov and Xpio-Tov (for 
reading (1) could mean 'the knowledge of the mystery of God 
Christ', or 'the knowledge of the mystery of God's Christ', or 
'the knowledge of God's mystery, of Christ'). Besides the inser- 
tion of TraTpos (readings 9-15) several other attempts were made 
to explain the relationship of XpiaTov to deov (4, 5, 6, 7, 8). The 
scribes responsible for readings (2) and (3) sought to relieve the 
difficulty by the elimination of one or the other of the two 
genitives — and in support! of (3) the scribe could point to Eph. 
iii. 4 as a precedent {tu> [ivaT-qpio) tov XpiaTov). Reading (4) 
gives what must be the right sense, suggesting that in reading ( 1 ) 

238 The Practice of New Testament Textual Criticism 

the word Xpiarov is explanatory of tov ixvarr^piov tov Oeov. 
Perhaps in dictating the Epistle the author separated the word 
XpioTov from the preceding phrase by a slight pause for breath, 
which can be represented in modern printing by a comma. 
Thus it is possible to explain the origin of all the other readings 
on the assumption that reading (1) is original, but this reading 
cannot be explained as derivative from any of them. Since the 
external support of (1) is of the best, both as regards age and 
character, one must conclude that tov Beov Xpiarov is the earliest 
attainable form of text preserved among the extant witnesses. 1 

As was remarked earlier, the harmonization of the Evan- 
gelists, whether done by scribes deliberately or unconsciously, 
is by definition a secondary process. Therefore the supreme rule 
for editors of the text is to give each Gospel its own proper 
character. Of many, many examples which might be cited as 
illustrations of this basic principle, the following has been 
chosen in order to show that on very rare occasions the correct 
reading may be preserved alone in the Koine or Byzantine text. 

If one were to take into consideration only the external 
evidence in support of dpxiepeis in Luke xx. 1 ('the chief 
priests and the scribes with the elders came up ...'), there would 
appear to be no doubt whatever concerning its originality — the 
diversity and the antiquity of the witnesses are most impressive, 
and the support for the variant reading UpeTs appears to be 
negligible. Thus: 

(1) dpxiepeis XBCDLMNQ,R@'F, fam. 1, fam. 13, 

fam. 1424, 33 157 579 892 107 1 1604, Old Lat, Vulg, 
Syr, Sah, Boh, Arm, Geo, Eth, Diatess Arab - Ital <ven), 

Dutch (Lifcge) 

(2) UpeTs AEGHKSUVWrJ/177, many minuscules, 


When one looks at a harmony of the Gospels, however, he 
discovers that the parallels to this account (Matt. xxi. 23 and 
Mark xi. 27) refer to 'the chief priests' in company with scribes 
and/or elders, with no fluctuation in the manuscript evidence. 

1 The caution with which the conclusion is expressed is in deference to those who, 
like F. W. Beare in the Interpreter's Bible, ad loc, think that the difficult expression 
toC 8c ov Xpiarov is too difficult for the author to have written, and that some primi- 
tive error must lie behind the earliest attainable text. 

The Practice of New Testament Textual Criticism 239 

Furthermore, a concordance provides the information that the 
same kind of stereotyped formula appears nearly fifty times 
throughout the Gospels and Acts, and in no case has lepeTs re- 
placed dpxiepeis. 

In the passage under consideration there is no discernible 
motive for altering 'chief priests' to 'priests', whereas the in- 
fluence of the stereotyped phrase containing apxiepeTs would 
have been felt independently by scribes of manuscripts in every 
textual tradition. One must conclude, therefore, that Teschen- 
dorf and von Soden are correct in printing lepets in Luke xx. 1, 
and that Westcott-Hort, Weiss, Vogels, Merk, Bover, Nestle, 
and the British and Foreign Bible Society's text of 1958 are 
wrong in preferring apxiepeLs. 1 

There are other instances besides Luke xx. 1 where it has 
happened that almost all of the 'good' manuscripts are in error 
and the correct reading is preserved in 'inferior' witnesses. Thus 
at Matt. xii. 47 the best Alexandrian witnesses (B X*), joining 
with representatives of the Western group (ff l kSyr s ' c ), are con- 
victed of having accidentally omitted a whole verse. Verse 47 
closes with the same word as verse 46, and the eyes of copyists 
have chanced to wander from the end of one to the end of the 

In Heb. vii. 1 the relative pronoun os, which is not suited to 
the syntax of the sentence, is supported by X A B D p I K p 33 al. 
The grammatically correct 6 is witnessed by p+ 6 C* and the 
Koine group of manuscripts. In this case one can see how the 
primitive error entered both the Alexandrian and Western 
traditions, for the following word begins with the letter sigma 
(awavTrjcras) , and so the correct o came to be pronounced and 
spelled 6'?. 

Occasionally considerations relating to intrinsic evidence will 
cast a decisive vote in the face of what appears to be over- 
whelming external testimony. The generally received text of 
Matt. xxii. 34-35 reads 'But when the Pharisees heard that he 
[Jesus] had silenced the Sadducees, they came together. And 

1 For another discussion of this passage, see Heinrich Greeven's comments in 
New Testament Studies, vi (i960), pp. 295 f., where he comes to the same conclusion 
as that adopted here. 

240 The Practice of New Testament Textual Criticism 

one of them, a lawyer, asked him a question, to test him.' This all 
seems straightforward enough and quite unexceptional. There 
are, however, a few witnesses which omit vopiKos ('a lawyer') at 
verse 35, namely fam. 1 , e, Syr s , Arm, Geo, and Origen accord- 
ing to Rufinus' Latin translation. Normally this evidence, 
comprising only part of the Caesarean text with some versional 
support, would not be regarded as especially weighty. In 
this case, however, internal considerations of the most compel- 
ling kind make it altogether likely that all of the uncials and 
almost all the minuscules are wrong in reading vopiKos. In the 
first place, except for this passage, it is Luke alone of the four 
Evangelists who employs the word vopiKos; the usual word in 
Matthew for one concerned with the Mosaic law is ypa/x/xctTev'?, 
a word which appears frequently in all three Synoptic Gospels. 
In the second place, an examination of a harmony of the Gospels 
discloses that in the parallel passage Luke reads 'And behold, a 
lawyer stood up to put him [Jesus] to the test. . .' (x. 25). 
There is no reason why scribes should have omitted vopwos 
from Matt. xxii. 35, but there would have been a strong 
temptation to add such a circumstantial detail from the parallel 
account in Luke. Despite, therefore, the rather limited range of 
evidence in support of the shorter text, it appears that to/hiko? is 
an intrusion into the text of Matt. xxii. 35. 

In 1 Thess. iii. 2, 'We sent Timothy, our brother and ... in 
the gospel of Christ, to establish you in your faith and to exhort 
you', the word or words after 'our brother and' have been trans- 
mitted in five different forms. They are the following : 

(1) 'God's fellow worker' (owepyov tov Beov, D p * 33 d e, Pela- 
gius, Ambrosiaster, pseudo-Hieronymus) . 

(2) 'fellow worker' (owepyov, B, Ephraem (as it seems)). 1 

(3) 'God's servant' (Siolkovov tov 9eo€, X A P p 424**, Goth, 
Boh, Arm, Syr h , Eth, Basil, Theod. Mops. lat ). 

(4) 'servant and God's fellow worker' (Slolkovov kclI ovvepyov 
tov 9eov, FpGp). 

(5) 'God's servant and our fellow worker' (Siclkovov tov 6eov 
Kal awepyov rjpa>v, D p(c) K p L p Syr p , most minuscules, 
Chrysostom, Theodoret). 

1 See Joseph Molitor, Der Paulustext des hi. Ephram aus seinem armenisch erhaltenen 
Paulinenkommentar . . . (Rome, 1938), p. 112. 

The Practice of Mew Testament Textual Criticism 241 

The reading placed last in the list, though supported by the 
great majority of witnesses, is obviously a secondary reading 
that was formed by joining readings (3) and (2) with the addi- 
tion of the word ij/xtov. This conflate reading, so typical of the 
Koine type of text, lies behind the rendering of the verse in the 
King James version. 

The fourth reading is also a type of conflated reading. The 
reviser has rather mechanically combined readings (3) and (1), 
writing the qualifying genitive tov deov only once. The external 
support for this variant is of little value, for the two bilingual 
manuscripts F and G are so closely related to each other as to 
constitute but one witness, and this witness belongs to a second- 
ary stratum of the Western type of text. We are left therefore with 
readings (1), (2), and (3) ; which of these gave rise to the others? 
It is difficult to answer this question with assurance, for, as it 
happens, more or less compelling reasons can be found for 
preferring each of the three readings as the original — reasons 
which led Bover and the translators of the New English Bible to 
adopt (1), Weiss to adopt (2), and Tischendorf, Westcott-Hort, 
von Soden, Vogels, Merk, and the Revised Standard Version 
to adopt (3). Arguments which support each of these choices 
involve the following considerations. 

The second reading, awepyov, can scarcely be original, for 
then one would have no adequate explanation for the addition 
of tov 6eov, producing reading (1), or its alteration to Skikovov 
with the addition of tov deov, producing reading (3) . 

It cannot be disputed that the third reading, 'God's servant', 
is attested by several excellent Alexandrian witnesses as well 
as by representatives of widely diversified versions. But the 
expression is both so clear and so rich that, if original, no one 
would have felt the need of altering it. The same cannot be 
said of the first reading in the list, for the application to Timothy 
of the expression 'God's fellow worker' may have offended the 
scrupulous. How can any man (so some may have reasoned) be 
given so exalted a title, least of all a young convert who was 
merely assisting Paul? Two courses were open to those who 
wished to soften the idea: by suppressing tou 6eov, Timothy 
becomes Paul's fellow worker (reading 2), in accord with his 
designation in Rom. xvi. 21; or by keeping tov deov and sub- 
stituting Skikovov for owepyov, Timothy is again described 

242 The Practice of New Testament Textual Criticism 

(reading 3) in typically Pauline language (Rom. xiii. 4; 2 Cor. 
vi. 4). Thus it would appear that reading (1) awepyov tov deov 
is original, for it explains the origin of the others. 

On the other hand, however, one's confidence in the pre- 
ceding arguments may well be weakened by a second review of 
the evidence. Those who are moved chiefly by considerations 
based on external evidence will certainly not be impressed by 
the combination of D p *, 33, d, e, Pelagius, Ambrosiaster, and 
pseudo-Hieronymus. Furthermore, the argument which was 
used above to discredit the originality of the reading awepyov tov 
deov, namely that it would have offended the scrupulous, be- 
comes less cogent when one observes that a similar expression 
in 1 Cor. iii. 9 (deov ydp ea/iev owepyol) has occasioned no textual 
variation in any known witness. Furthermore, the generally ac- 
knowledged tendency of Western texts to substitute synonyms 
and near-synonyms suggests that reading ( 1 ) arose as a scribal 
alteration of (3). 

The chief weakness, however, of the argument in favour of 
the originality of reading (3) is the absence of any good reason 
to account for the deletion of tov deov to produce the second 
reading. Weiss therefore argues that the simplest explanation to 
account for the rise of the many forms of the text is to regard 
awepyov of B as the original, of which all the others are modi- 
fications. 1 

In making an assessment of the weight of the argument sup- 
porting each of the variant readings, one will do well to exercise 
great caution. According to Lachmann the degree of certitude 
or probability attaching to any particular variant may be 
assessed as certain, almost certain, probable, doubtful, very 
doubtful, or nil. In this case, whatever choice is made among the 
three chief variants, every candid critic must acknowledge the 
strength of the arguments supporting the other two. 

Another example where violent conflict between external 
and internal evidence makes it impossible to come to a con- 
fident decision among variant readings is found in Acts xii. 25. 
The natural impression which one gets when reading the section 
xi. 27 to xiii. 1 is that xi. 30 refers to the arrival of Paul and 

1 B. Weiss, Textkritik der paulinischen Briefe (= Texte und Untersuchungen, xiv (3) ; 
Leipzig, 1896), pp. 13 f. It is significant that in this case Westcott and Hort did not 
see fit to follow their favourite manuscript. 

The Practice of Mew Testament Textual Criticism 243 

Barnabas at Jerusalem and that xii. 25 ought to tell of their 
departure from Jerusalem. There is, however, early and rather 
widespread support for the almost impossible reading «'? 'Iepov- 
oa\-qii (X B H a L a P a 81, &c, Syr hm «, 1 Sah MS., 1 Eth r °). The 
other readings are ££ 'IepovoaXruj. (A 33 547 913 1739 1891, Sah, 
Boh, Arm) and ano 'hpovaaXrjfx (D 181 431 614, &c, gig, Vulg; 
Syr p witnesses to either i£ or 6.tt6) . 

All the canons of textual criticism favour the more difficult 
reading «'?, supported as it is by the earliest and best manu- 
scripts; but the sense of the context cries out for a preposition 
that means 'from'. 2 Was «? a primitive corruption which later 
scribes attempted to correct by altering it to i$ or a-n-d ? Here one 
must acknowledge that he simply does not know what the author 
originally wrote. 3 

Does Luke say that Jesus sent out seventy or seventy-two 
disciples (Luke x. 1, 17)? The evidence is as follows: 

ePSo/MTJKovTa is supported by p 45 (extant for vs. 17 only) 4 
XACLW/J & A S, fam. 1 , fam. 1 3, most minuscules, f q i 

1 This Sahidic manuscript is B.M. Or. 7594, thought by some to date from the 
fourth century; see n. 1 on p. 79 above. 

2 According to the addenda to Tischendorf's 8th ed. (as reported in Gregory's 
Prolegomena, p. 1 28 1 ) , the scribe of B had begun to write A.tto, but then changed it to 

3 Hort (Appendix, p. 94) suggested that the passage be emended by transpos- 
ing the order of the words, thus separating els 'IepovaaXrjp. from vTrearpei/iav and 
ranging it with what follows (vire<Trpeipav ttjv els ' ' lepovaaXrfp. TrX-qpwaavTes StaKOvlav, 
'they returned, having fulfilled their ministry at Jerusalem'). More recently DomJ. 
Dupont concluded a lengthy examination of the verse by suggesting that the pre- 
position els be retained and the difficulty in sense be alleviated by punctuating 
with a comma after the main verb (virecTpeiftav, els * lepovaaX-fjp: TrXripuiaavTes rrjv 
SmKoia'av) ; see his article, 'La Mission de Paul "a Jerusalem" (Actes xii. 25)', 
Novum Testamentum, i (1956), pp. 275-303. 

4 F. G. Kenyon erroneously reported the evidence of p 45 as supporting the 
numeral 72 (o/3). The present writer has examined this passage in J) 45 under natural 
and artificial light, and has assured himself that the Greek character which follows 
the omicron is nothing but a diple, or space-filler ( >), which scribes used occasion- 
ally in order to bring an otherwise short line even with the right-hand margin of the 
column. In fact, by consulting Kenyon's volume of plates of p 45 anyone can see the 
similarity between the disputed character and the diple which appears on the same 
folio near the top of the column. For a fuller discussion of the reading of p 4s , in- 
cluding C. H. Roberts's view that the numeral is 76 (o?), see B. M. Metzger, 
'Seventy or Seventy-two Disciples?', in Mew Testament Studies, v (1959), pp. 299- 
306, on the basis of which the citation of the evidence of p 4s was corrected in the 
i960 edition of Nestle's Greek Testament. Other editions of the Greek Testament 
continue to repeat the error in Kenyon's edition of p 4s . 

244 The Practice of New Testament Textual Criticism 

(extant for vs. i only), Syr° (vs - "'■ p- h 'P al , Cop boh , Goth, Iren, 
Tert, Clem, Orig, Eus, Ambr, Jerome. 

e^So^Kovra bvo is supported by p 75 B D M (vs. I only) 0181 
(vs. 1 only) abcdlr 2 , Vulg, Syr s > c (vs - °' h m s (vs - "> , Cop sah , Georg 1 - 2 , 
Diatess Dutch Pers Ital , Orig, Ephr, Aug, Ambrst. 

The external evidence is almost evenly divided; the chief 
representatives of the Alexandrian and the Western groups, with 
most of the Old Latin and the Sinaitic Syriac, support the 
numeral seventy-two. On the other hand, other Alexandrian evi- 
dence of relatively great weight (HL.AAS) as well as Caesarean 
witnesses (p 45 , fam. 1, fam. 13) join in support of the numeral 

The factors bearing upon the evaluation of internal evidence, 
whether involving transcriptional or intrinsic probabilities, are 
singularly elusive. It is likely that in most of the early manuscripts 
(as in p 45 and p 75 ) the numeral was written with letters of the 
alphabet (either o/J or 5). It was easy, therefore, for either 
number to be accidentally altered to the other. If the variation was 
made deliberately, one can imagine that an Alexandrian scribe 
with a mathematical penchant altered seventy to seventy-two 
for the sake of what he may have regarded as scholarly symmetry. 
On the other hand, if the alteration was made unwittingly, it 
is perhaps more likely that the precise number should be trans- 
formed into the round number seventy than that the 'solemn' 
number seventy should be transformed into seventy-two. 

Those who transmitted the account prior to its inclusion in 
Luke x may have wished to convey a symbolic meaning in the 
number of the disciples, and it is easy to find parallels in Jewish 
antiquities for either seventy or seventy-two. Seventy elders were 
appointed by Moses to assist him (Num. xi. 16-17, 24-25) ; there 
were seventy sons of Jerubbaal (Judges ix. 2) , seventy sons of Ahab 
(2 Kings x. 1), and seventy priests of Bel (Bel and Dragon, vs. 10). 
On the other hand, according to the Letter of Aristeas 
(§§ 46-50) seventy-two elders (six from each of the twelve tribes) 
were chosen in order to prepare a Greek translation of the 
Torah (the Septuagint), and in 3 Enoch the number of princes 
of kingdoms on high is seventy-two, corresponding to the 
seventy-two languages of the world (xvii. 8 ; cf. xviii. 2 f., xxx. 2) .' 

1 For many other examples from Jewish antiquities involving seventy and 
seventy-two, see the article referred to in the previous footnote. 

The Practice of New Testament Textual Criticism 245 

It is, however, exceedingly difficult to ascertain what sym- 
bolism is intended in Luke's account. On the one hand, if the 
mission of this group of disciples is to be understood as a mission 
to Israel, the number may have been chosen as a multiple 
of the twelve tribes of Israel. On the other hand, since several 
New Testament writers presuppose a parallel between Jesus 
and Moses, 1 it may be that this group of Jesus' disciples is in- 
tended to correspond to the seventy elders who assisted Moses. So 
evenly balanced are these two possibilities that it is hazardous 
to dogmatize as to which is more probable. 

A total appraisal of both external and internal evidence 
bearing on these variant readings must remain indecisive. 
Though the reading 'seventy-two' is supported by a combina- 
tion of early witnesses that normally carries a high degree of 
conviction of originality, yet the diversity of witnesses which 
read 'seventy' is so weighty and internal considerations are so 
evenly balanced that the textual critic must simply acknow- 
ledge his inability to decide with assurance between the two. If 
one is editing the Greek text of Luke perhaps the least un- 
satisfactory resolution of the dilemma is to have recourse to 
brackets (which are always a tacit confession of the editor's 
uncertainty) and to print i^ho^Kovra [8vo]. 

The passages discussed above will suffice as specimens illustrat- 
ing the wide variety of considerations which one must take into 
account when evaluating variant readings. In addition to solv- 
ing such problems individually as they occur in the text, an 
editor of the' Greek Testament is confronted with still others 
arising from the diversity of usage and authorship within the 
New Testament as a whole. How far, for example, should he 
attempt to standardize first aorist endings on second aorist 
stems? Shall he reproduce in his edition divergencies between 
two or more authors in the spelling of the same proper name? 
Or, even within the same book, is it permissible to* vary the 
spelling of a proper name? For example, shall he print the 
title of Paul's Epistle to the Colossians TIPOZ KOAAEEAEIE 
(with p+ 6 A B* I K p and many manuscripts), as the Textus Re- 
ceptus, Lachmann, and Westcott and Hort do, but spell the 

1 See J. Jeremias in Kittel, Theologisches Worterbuch zum Neuen Testament, iv, 
pp. 871-8. 

246 The Practice of New Testament Textual Criticism 

name of the city in Col. i. 2 KoXoooals (with p+ 6 X [A has a 
lacuna] B D p E? Fp G? U>, Sac.) ? It will be understood that it is 
far easier to ask such questions than to answer them. 

By way of conclusion, let it be emphasized again that no 
single manuscript and no one group of manuscripts exists which 
the textual critic may follow mechanically. All known witnesses 
of the New Testament are to a greater or less extent mixed texts, 
and even the earliest manuscripts are not free from egregious 
errors. Although in very many cases the textual critic is able to 
ascertain without residual doubt which reading must have 
stood in the original, there are not a few other cases where he 
can come only to a tentative decision based on an equivocal 
balancing of probabilities. Occasionally none of the variant 
readings will commend itself as original, and he will be com- 
pelled either to choose the reading which he judges to be the 
least unsatisfactory or to indulge in conjectural emendation. In 
textual criticism, as in other areas of historical research, one 
must seek not only to learn what can be known, but also to be- 
come aware of what, because of conflicting witnesses, cannot 
be known. 


Check-list of the Greek Papyri of the 
New Testament 

The following list supplies a brief conspectus of basic information 
concerning (a) the contents of each papyrus document of the Greek 
New Testament; {b) the latest opinion regarding its approximate 
date, 1 given by century in Roman numerals; {c) the present location 
of the papyrus ; {d) the bibliographical reference to the editio princeps 
of the papyrus; and (e) the text- type or family to which the papyrus 
has been thought to belong. (In some cases too little text has been pre- 
served to warrant making a judgement concerning textual affinities.) 
All of the papyri are in codex-form except p' 2 ,p 13 , p 18 , and p 43 (?) 
which are fragments from rolls. (See below, p. 273.) 


P.Oxy. = B. P. Grenfell and A. S. Hunt, The Oxyrhynchus Papyri, 

i- (London, 1898-). 
P.S.I. = Pubblitazioni della Societd Italiana, Papiri Greci e Latini. 
Sanz = Peter Sanz, Griechische literarische Papyri christlichen Inhalts, 

i (Vienna, 1946). 

p 1 (a) Matt. i. 1-9, 12, 14-20, 23; (b) ill; (c) Philadelphia, Uni- 
versity of Pennsylvania Museum, no. E 2746; {d) P.Oxy. i, 
pp. 4-7, no. 2; (e) Alexandrian text. 

p 2 (a) Johnxii. 12-15; (b) vi; (c) Florence, Museo Archeologico, 
Inv. no. 7134; (d) E. Pistelli in Studi religiosi, vi (1906), 
pp. 129-30; («) mixed text. 

p 3 (a) Luke vii. 36-45 ; x. 38-42 ; (b) vi/vn ; (c) Vienna, Osterreichi- 
sche Nationalbibliothek, Sammlung Papyrus Erzherzog Rainer, 
no.G2323; (d) K. Wessely in Wiener Studien, iv (1882), pp. 198- 
9, 211-14; vii (1885), PP- 69-70; and J. N. Birdsall, ibid., lxxvi 
(1963), pp. 163-4; W Alexandrian text. 

1 The dates are those suggested by the following palaeographers who based 
their judgements upon a comprehensive evaluation of all the New Testament 
Greek papyri: Herbert Hunger of Vienna, Karl Preisendanz of Heidelberg, C. H. 
Roberts of Oxford, the lateWilhelm Schubart of Halle, and T. C. Skeat of London 
(see Kurt Aland, New Testament Studies, ix [1963], pp. 303-16). 

248 Appendix 

p+ (a) Luke i. 58-59, 62-ii. 1,6-7; &■ 8-38; iv. 2, 29-32, 34~355 
v. 3-8, 30-38; vi. 1-16; (b) in; (c) Paris, Bibliotheque Nationale, 
no. Gr. 1 120, suppl. 2 ; (d) M.-J. Lagrange, Critique textuelle, 
pp. 1 19-23; cf. Jean Merell, Revue biblique, xlvii (1938), pp. 5-22 ; 
(e) Alexandrian text. 

9 s (a) John i. 23-31, 33-41; xvi. 14-30; xx. 11-17, 19-20, 22-25; 

(b) 111; (c) London, British Museum, P. 782 and P. 2484; 

(d) P.Oxy. ii, pp. 1-8, no. 208, and P.Oxy. xv, pp. 8-12, 
no. 1 781; (e) Western text. 

p 6 (a) John x. 1-2, 4-7, 9-10; xi. 1-8, 45-52; (A) rv; 

(c) Strasbourg, Bibliotheque de la University, 35 i r , 335 v , 379, 
38 1 , 383, 384 copt. ; (d) F. Rosch, Bruchstucke des I. Clemensbriefes 
nach dem achmimischen Papyrus der Strassburger Universitdts- und 
Landesbibliothek (Strassburg, 1910), pp. 119-22, 131-4, 143-8; 

(e) agrees with B and 0. 

p 7 (a) Luke iv. 1-2; (b) v; (c) now lost; was in Kiev, library of the 
Ukrainian Academy of Sciences; (d) C. R. Gregory, Textkritik 
des Neuen Testamentes, iii, p. 1086, and K. Aland in New Testa- 
ment Studies, iii (1957), pp. 262-4. 

p 8 (a) Acts iv. 31-37; v. 2-9; vi. 1-6, 8-15; (b) rv; (c) now 
lost; was in Berlin, Staatliche Museen, P. 8683; (d) C. R. 
Gregory, Textkritik des Neuen Testamentes, iii, pp. 1086-90; 
(e) mixture of Alexandrian and Western texts. 

p' (a) 1 John iv. 11-12, 14-17; (b) in; (c) Cambridge, Massa- 
chusetts, Harvard University, Semitic Museum, no. 3736; 
(</) P.Oxy. iii, pp. 2-3, no. 402. 

p 10 (a) Rom. i. 1-7; (b) iv; (c) Cambridge, Massachusetts, Harvard 
University, Semitic Museum, no. 2218; (d) P.Oxy. ii, pp. 8-9, 
no. 209; («) Alexandrian text. 

p 11 (a) 1 Cor. i. 17-23; ii. 9-12, 14; iii. 1-3, 5-6; iv. 3-y. 5, 7-8; 
vi. 5-7, 11-18; vii. 3-6, 10-14; (b) vn; (c) Leningrad, State 
Public Library; (d) Tischendorf in Verhandlungen der 25. Ver- 
sammlung der deutschen Philologen und Schulmanner in Halle (Leipzig, 
1868), pp. 44-45, and K. Aland in New Testament Studies, iii 
(1957), pp. 268-78; (e) Alexandrian text. 

p 12 (a) Heb. i. 1; (b) end of in; (c) New York, Pierpont Morgan 
Library, no. G. 3 ; (d) Grenfell and Hunt in The Amherst Papyri, i 
(London, 1900), pp. 30-31. 

p 13 (a) Heb. ii. 14-v. 5; x. 8-22, 29-xii. 17; (b) m/iy (perhaps 
first half of fourth century) ; (c) London, British Museum, 
P. 1 532 (verso) , and Florence, Biblioteca Medicea Laurenziana ; 

Appendix 249 

(d) P.Oxy.iv, pp. 36-48,110. 657; andP.S.I. xii (1951), pp. 209- 
10, no. 1292; («) Alexandrian text. 

j 1 * (a) 1 Cor. i. 25-27; ii. 6-8; iii. 8-10, 20; (b) v (?) ; (c) Mount 
Sinai, St. Catharine's Monastery, no. 14; (d) J. Rendel Harris, 
Biblical Fragments from Mount Sinai (London, 1890), pp. xiii, 
54-56; (e) Alexandrian text. 

p's (a) 1 Cor. vii. 18-viii. 4; (b) in; (c) Cairo, Museum of 
Antiquities, no. 47423; (d) P.Oxy. vii, pp. 4-8, no. 1008; 

(e) Alexandrian text. 

(C* (a) Phil. iii. 9-17; iv. 2-8; (b) m/iv; (c) Cairo, Museum of 
Antiquities, no. 47424; (d) P.Oxy. vii, pp. 8-1 1, no. 1009; 
(e) Alexandrian text. 

p 17 (a) Heb. ix. 12-19; (b) iv; (c) Cambridge, England, University 
Library, gr. theol. f. 13 (P), Add. 5893; (d) P.Oxy. viii, pp. 1 1- 
13, no. 1078; (e) mixed text. 

p 18 (a) Rev. i. 4-7; (b) m/iv; (c) London, British Museum, P. 2053 
(verso); (d) P.Oxy. viii, pp. 13-14, no. 1079; («) agrees with 
X, A, and C. 

p'» (a) Matt. x. 32-xi. 5 ; (b) rv/v; (c) Oxford, Bodleian Library, MS. 
Gr. bibl. d. 6 (P.) ; (d) P.Oxy. ix, pp. 7-9, no. 1 170; (e) mixed 
text, chiefly Western. 

P 20 ( a ) J as - n - z 9 _m - 2 > 4—9 ; (b) in; (c) Princeton, New Jersey, Uni- 
versity Library, Classical Seminary AM 41 17 (15) ; (d) P.Oxy. 
ix, pp. 9-1 1, no. 1 171 ; (e) Alexandrian text. 

p" (a) Matt. xii. 24-26, 31-33; (b) rv/v; (c) Allentown, Penn- 
sylvania, Library of Muhlenberg College, Theol. pap. 3; 

(d) P.Oxy. x, pp. 12-14, no. 1227; (e) agrees with D and X corr . 

P" (a) John xv. 25-27; xvi. 1-2, 21-32; (b) in; (c) Glasgow, Uni- 
versity Library, MS. 2-x. 1 ; [d) P.Oxy. x, pp. 14-16, no. 1228; 

(e) mixed text, agrees partly with H«>" an d partly with D. 

P 23 (a) Jas. i. 10-12, 15-18; (b) beginning of hi; (c) Urbana, Illinois, 
University of Illinois, Classical Archaeological and Art Museum, 
G. P. 1229; (d) P.Oxy. x, pp. 16-18, no. 1229; (e) Alexandrian. 

p 24 (a) Rev. v. 5-8; vi. 5-8; (b) iv; (c) Newton Center, Massa- 
chusetts, Library of Andover Newton Theological School; 
(d) P.Oxy. x, pp. 18-19, no. 1230; (e) agrees with X. 

p 25 (a) Matt, xviii. 32-34; xix. 1-3, 5-7, 9-10; (b) end of iv; 

(c) now lost; was in Berlin, Staatliche Museen, P. 16388; 

(d) Otto Stegmiiller in Jfjitschriftfur die neutestamentliche Wissen- 
schaft, xxxvii (1938), pp. 223-9; (e) Western text. 

250 Appendix 

p 26 (a) Rom. i. 1-16; (Jb) c. 600; (c) Dallas, Texas, Southern 
Methodist University, Lane Museum; (<f) P.Oxy. xi, pp. 6-9, 
no. 1354; (e) agrees with A and X. 

p« (a) Rom. viii. 12-22, 24-27, 33-39; ix - '-3. 5~9; (*) IU 1 

(c) Cambridge, England, University Library, Add. MS. 7211; 
(<f) P.Oxy. xi, pp. 9-12, no. 1355; (e) chiefly Alexandrian text, 
but with Western readings. 

p 28 (a) John vi. 8-12, 17-22; (b) in; (c) Berkeley, California, 
Library of Pacific School of Religion, Pap. 2 ; (d) P.Oxy. xiii, 
pp. 8-10, no. 1596; (e) Alexandrian text. 

p 2 » (a) Acts xxvi. 7-8, 20; (b) in; (c) Oxford, Bodleian Library, 
MS. Gr. bibl. g. 4 (P.); (<f) P.Oxy. xiii, pp. 10-12, no. 1597; 
(e) Western text (?). 

p 30 (a) 1 Thess. iv. 13, 16-18; v. 3, 8-10, 12-18, 26-28; 2 Thess. 
i. 1-2; (b) in; (c) Ghent, University Library, U. Lib. P. 61; 

(d) P.Oxy. xiii, pp. 12-14, no. 1598; (e) mixed text. 

p 31 (a) Rom. xii. 3-8; (b) vn; (c) Manchester, England, John 
Rylands Library, P. Ryl. 4; (<f) A. S. Hunt, Catalogue of Greek 
Papyri in the John Rylands Library, i (Manchester, 191 1), p. 9; 
(«) agrees with X. 

p 32 (a) Titus i. 1 1— 15; ii. 3-8; (b) c. 200; (c) Manchester, England, 
John Rylands Library, P. Ryl. 5; (d) A. S. Hunt, Catalogue of 
Greek Papyri in the John Rylands Library, i (Manchester, 191 1), 
pp. 10-1 1 ; (e) agrees with X, also with F and G. 

p 33 (a) Acts xv. 22-24, 27-32; (b) vi ; (c) Vienna, Osterreichische 
Nationalbibliothek, no. 190; (d) C. Wessely, Studien zur 
Palaeographie und Papyruskunde, xii (19 14), p. 245; («) chiefly 
Alexandrian text. 

p 34 (a) 1 Cor. xvi. 4-7, 10; 2 Cor. v. 18-21; x. 13-14; xi. 2, 4, 
6-7; (b) vn ; (c) Vienna, Osterreichische Nationalbibliothek, 
no. 191; (d) C. Wessely, Studien zur Palaeographie und Papyrus- 
kunde, xii (1914), p. 246; (e) Alexandrian text. 

p 3S (a) Matt. xxv. 12-15, 20-23 ; (b) iv (?) ; (c) Florence, Biblioteca 
Medicea Laurenziana; (d) P.S.I, i (1912), pp. 1-2; (e) mixed 
text (Alexandrian and Western). 

p 36 (a) John iii. 14-18, 31-32; (b) vi; (c) Florence, Biblioteca 
Medicea Laurenziana; (d) P.S.I, i (1912), pp. 5-6; (e) mixed 
text (Western and Alexandrian). 

Appendix 251 

p 37 (a) Matt. xxvi. 19-52; (b) iii/iv; (c) Ann Arbor, Michigan, 
University of Michigan Library, Invent, no. 1570; (d) H. A. 
Sanders in Harvard Theological Review, xix (1926), pp. 215-26, 
and in Michigan Papyri, pp. 9-14; (e) Caesarean text. 

p 38 (a) Acts xviii. 27-xix. 6, 12-16; (b) c. 300; (c) Ann Arbor, 
Michigan, University of Michigan Library, Invent, no. 1571 ; 

(d) H. A. Sanders in Harvard Theological Review, xx (1927), 
pp. 1-19, and in Michigan Papyri, pp. 14-19; (e) Western text. 

p 39 (a) John viii. 14-22; (b) m; (c) Chester, Pennsylvania, Crozer 
Theological Seminary Library, no. 8864; (d) P.Oxy. xv, pp. 7- 
8, no. 1 780 ; (e) Alexandrian text. 

p*° (a) Rom. i. 24-27, 3 1 — ii- 3; iii. 21-iv. 8; vi. 4-5, 16; ix. 17, 
27; (b) 111; (c) Heidelberg, Universitatsbibliothek, Inv. Pap. 
graec. 45; (d) F. Bilabel, Griechische Papyri (Heidelberg, 1924), 
pp. 28-31 (= Verqffentlichungen aus den Badischen Papyrus- 
Sammlungen, iv, pp. 124-7); ( e ) Alexandrian text. 

p 41 (a) Acts xvii. 28-xviii. 2, 24-25, 27; xix. 1-4, 6-8, 13-16, 18- 
19; xx. 9-13, 15-16, 22-24, 26-28, 35-38; xxi. 1-4; xxii. 11- 
14, 16-17; W VIII J ( c ) Vienna, Osterreichische Nationalbiblio- 
thek, Pap. K. 7541-8 ; (d) C. Wessely, Studien zur Palaeographie und 
Papyruskunde, xv (1914), pp. 107-18; (e) Western text. 

p 42 (a) Luke i. 54-55; ii. 29-32; (b) vii/vm; (c) Vienna, Osterreichi- 
sche Nationalbibliothek, KG 8706; (d) Walter Till and Peter 
Sanz, Eine griechisch-koptische Odenhandschrift (=Monumenta biblica 
et ecclesiastica, v; Rome, 1939), p. 112; (e) agrees with A. 

p 43 (a) Rev. ii. 12-13; xv. 8-xvi. 2; (b) vi/vii; (c) London, British 
Museum, Pap. 2241 ; (c) W. E. Crum and H. I. Bell, Wadi 
Sarga: Coptic and Greek Texts (Copenhagen, 1922), pp. 43-45; 

(e) chiefly Alexandrian text. 

p 44 (a) Matt. xvii. 1-3, 6-7; xviii. 15-17, 19; xxv. 8-10; John 
x. 8-14; ix. 3-4; xii. 16-18 [in this order]; (b) vi/vii; (c) New 
York, Metropolitan Museum of Art, Inv. 14-1-527; (d) H. G. 
Evelyn White, The Monastery of Epiphanius at Thebes, ii (New 
York, 1926), pp. 120-1, 301; (e) Alexandrian text. 

p 4S (a) Matt. xx. 24-32; xxi. 13-19; xxv. 41-46; xxvi. 1-39; Mark 
iv. 36-40; v. 15-26, 38-vi. 3, 16-25, 3 6 -5o; vii. 3-15, 25- 
viii. 1, 10-26, 34-ix. 8, 18-31; xi. 27-33; xii. i»5- 8 > l Z~ l 9> 
24-28; Luke vi. 31-41, 45-vii. 7; ix. 26-41, 45-x. 1, 6-22, 

252 Appendix 

26-xi. 1, 6-25, 28-46, 50-xii. 12, 18-37, 42-xiii. 1, 6-24, 29- 
xiv. 10, 17-33; J ohn x - 7-25. St-xi- I0 . 18-36, 43-57; Acts 
^.'27-36; v. 10-20, 30-39; vi. 7-vii. 2, 10-21, 32-41, 52-viii. 1, 
14-25. 34-i x - 6 . ^-27, 35-x. 2, 10-23, S 1 ^ 1 ; xi - 2-14. 24-xii. 
5, 13-22; xiii. 6-16, 25-36, 46-xiv. 3, 15-23; xv. 2-7, 19-26, 
38-xvi. 4, 15-21, 32-40; xvii. 9-17; (b) m; (c) Dublin, Chester 
Beatty Museum; and Vienna, Osterreichische Nationalbiblio- 
thek, P. Gr. Vind. 31974; (d)T. G. Kenyon, The Chester Beatty 
Biblical Papyri, fasc. ii, Gospels and Acts (London, 1933), and 
Hans Gerstinger in Aegyptus, xiii (1933), pp. 67-72; (e) partly 
Alexandrian, partly Western (pre-Caesarean) text. 

p* (a) Rom. v. 17-vi. 3, 5-14; viii. 15-25, 27-35, 37-«- 32; 
x. i-xi. 22, 24-33, 35-xiv. 8, 9-xv. 9 (fragmentary), n-33; 
xvi. 1-23, 25-27; Heb., 1 and 2 Cor., Eph., Gal., Phil., Col. 
(all with lacunae); 1 Thess. i. 1, 9-10; ii. 1-3; v. 5-9, 23-28; 

(b) c. 200; (c) Dublin, Chester Beatty Museum, and Ann Arbor, 
Michigan, University of Michigan Library, Invent, no. 6238; 

(d) F. G. Kenyon, The Chester Beatty Biblical Papyri, fasc. iii 
(London, 1934), and fasc. iii Supplement (London, 1936), and 
H. A. Sanders, A Third-century Papyrus Codex of the Epistles of Paul 
(Ann Arbor, 1935); (e) Alexandrian text. 

p 47 (a) Rev. ix. 10-xvii. 2 (with small lacunae); (b) end of in; 

(c) Dublin, Chester Beatty Museum; (d) F. G. Kenyon, The 
Chester Beatty Biblical Papyri, fasc. iii (London, 1934); («) agrees 
with A, C, and X. 

p 48 (a) Acts xxiii. 11-17, 23-29; (b) end of m; (c) Florence, Mu- 
seo Medicea Laurenziana; (d) P.S.I, x (1932), pp. 1 12-18; 

(e) Western text. 

p« (a) Eph. iv. 16-29, 3 I_V - r 3. W enc ^ of m; (c) New Haven, 
Connecticut, Yale University Library, P. 415; (d) W. H. P. 
Hatch and C. B. Welles, Harvard Theological Review, li (1958), 
pp. 33-35; (e) Alexandrian text. 

p so (a) Acts viii. 26-32; x. 26-31; (b) iv/v; (c) New Haven, Con- 
necticut, Yale University Library, P. 1543; (d) C. H. Kraeling 
in Quantulacumque, Festschrift for Kirsopp Lake (London, 
1937), pp. 163-72; (e) agrees chiefly with B. 

p' 1 (a) Gal. i. 2-10, 13, 16-20; (b) c. 400; (c) London, British 
Museum; (d) P.Oxy. xviii, pp. 1-3, no. 2157; (e) partly 
Alexandrian, partly eclectic text. 

Appendix , 253 

P 52 ( a ) J°hn xviii. 31-33, 37-38; (b) beginning of 11; (c) Manchester, 
John Rylands Library, P. Ryl. Gr. 457; {d) C. H. Roberts, 
An Unpublished Fragment of the Fourth Gospel in the John Rylands 
Library (Manchester, 1935); («) Alexandrian text. 

ps3 ( a ) Matt. xxvi. 29-40; Acts ix. 33-38, 40-x. 1 ; (b) in; (c) Ann 
Arbor, Michigan, University of Michigan Library, Invent, 
no. 6652; (d) H. A. Sanders in Quantulacumque, Festschrift for 
Kirsopp Lake (London, 1937), pp. 151-61 ; («) Egyptian mixed 

P s+ ( a ) J as - n - 16-18, 22-26; iii. 2-45(6) v/vi; (c) Princeton, New 
Jersey, Princeton University Library, Garrett Depos. 7742; 

(d) E. H. Kase, Jr., Papyri in the Princeton University Collections, ii 
(Princeton, 1936), pp. 1-3; (e) agrees with B, X, and C. 

p 5s (a) John i. 31-33, 35-38, on upper portion of page, hermeneia 
below (as p S9 and p 60 ); (b) vi/vn; (c) Vienna, Osterreichische 
Nationalbibliothek, P. Gr. Vind. 26214; (d) Sanz, pp. 58-59; 

(e) Alexandrian text. 

p 56 (a) Acts i. 1, 4-5, 7, 10-1 1 ; (b) v/vi; (c) Vienna, Osterreichische 
Nationalbibliothek, P. Gr. Vind. 19918; (d) Sanz, pp. 65-66; 
(e) Alexandrian text. 

p" (a) Acts iv. 36-v. 2, 8-10; (b) iv/v; (c) Vienna, Osterreichi- 
sche Nationalbibliothek, P. Gr. Vind. 26020; (d) Sanz, pp. 61- 
67; («) Alexandrian text. 

p* 8 (a) Acts vii. 6-10, 13-18; (b) vi; (c) Vienna, Osterreichische 
Nationalbibliothek, P. Gr. Vind. 17973, 36133 s4 , and 35831; 
{d) Sanz, pp. 67-68; («) Alexandrian, agreeing partly with 

p 59 {a) John i. 26, 28, 48, 51 ; ii. 15-16; xi. 40-52; xii. 25, 29, 31, 
35; xvii. 24-26; xviii. 1-2, 16-17, 22; xxi. 7, 12-13, 15, 17-20, 
23, on upper portion of page, hermeneia below (as p ss and p 60 ); 
(b) vn; (c) New York, New York University, Washington 
Square College of Arts and Sciences, Department of Classics, 
P. Colt. 3; (d) Lionel Casson and E. L. Hettich, Excavations 
at Nessana, ii (Princeton, 1950), pp. 79-93. 

P 60 ( a ) J°h n xvi. 29-xix. 26 with lacunae (probably contained 
hermeneia, as p 5S and p s ') ; (b) vn ; (c) New York, New York 
University, Washington Square College of Arts and Sciences, 
Department of Classics, P. Colt. 4; (d) Lionel Casson and E. L. 
Hettich, Excavations at Nessana, ii (Princeton, 1950), pp. 94-1 1 1 ; 
(e) Alexandrian text. 

254 Appendix 

p 61 (a) Rom. xvi. 23, 25-27; 1 Cor. i. 1-2, 4-6; v. 1-3, 5-6, 9-13; 
Phil. iii. 5-9, 12-16; Col. i. 3-7, 9-13; iv. 15; 1 Thess. i. 2- 
3; Titus iii. 1-5, 8-1 1, 14-15; Philem., vss. 4-7; (b) c. 700; 

(c) New York, New York University, Washington Square Col- 
lege of Arts and Sciences, Department of Classics, P. Colt. 5 ; 

(d) Lionel Casson and E. L. Hettich, Excavations at Nessana, 
ii (Princeton, 1950), pp. 112-22; (e) probably Egyptian. 

p 62 (a) Matt. xi. 25-30; (b) iv; (c) Oslo, University Library; 
(d) Leiv Amundsen in Symbolae Osloenses, xxiv (1945), pp. 121- 
40; (e) Alexandrian text. 

P 63 ( a ) J orm iii. 14-18; iv. 9-10; (b) c. 500; (c) Berlin, Staatliche 
Museen; (d) Otto Stegmiiller in Biblica, xxxiv (1953), 
pp. 13-22. 

p 64 (a) Matt. xxvi. 7, 10, 14-15, 22-23, 3 I_ 33> W c - 200 >' ( c ) Oxford, 
Magdalen College Library; {d) Colin Roberts in Harvard Theo- 
logical Review, xlvi (1953), pp. 233-7. 

p 6s (a) 1 Thess. i. 3-10; ii. 1, 6-13; (b) m; (c) Florence, Biblioteca 
Medicea Laurenziana; (d) P.S.I, xiv (1957), pp. 5-7; (e) Alex- 
andrian text. 

p 66 John i. i-vi. 11, 35-xiv. 26, and fragments of xiv. 29-xxi. 9; 
(b) c. 200; (c) Cologny/Geneve, Bibliotheque Bodmer; 
(d) V. Martin, Papyrus Bodmer II: £vangile de Jean, chs. 1-14 
(Bibliotheca Bodmeriana, 1 956) ; Supplement, £vangile de Jean, 
chs. 14-21 (Bibliotheca Bodmeriana, 1958); nouvelle ed., V. 
Martin et J. W. B. Barns (1962); (e) mixed text. 

p 67 Matt. iii. 9, 15; v. 20-22, 25-28; (b) c. 200; (c) Barcelona, Fun- 
dacion San Lucas Evangelista, P. Bare. 1 ; (d) P. Roca-Puig, 
Un papiro griego del evangelio de San Mateo (Barcelona, 1 956) ; 
2nd ed., with a Note by Colin [H.] Roberts ( 1 962) ; (e) agrees 
with X. 

p<* (a) 1 Cor. iv. 12-17, 19-21; v. 1-3; (b) vii(?); (c) Leningrad, 
State Public Library, Gr. 258; (d) K. Aland in New Testament 
Studies, iii (1957), pp. 266-9; (e) agrees with the Textus Re- 
ceptus against the Alexandrian text. 

p 6 » Luke xxii. 41, 45-48, 58-61; (b) in; (c) place (?); (d) P.Oxy. 
xxiv, pp. 1-4, no. 2383; (e) mixed text. 

p 70 (a) Matt. xi. 26-27; x "- 4~5J (*) ln 5 ( c ) place (?); (d) P.Oxy. 
xxiv, pp. 4-5, no. 2384. 

p 71 Matt. xix. 10- 11, 17-18; (b) iv; (c) place (?); (d) P.Oxy. xxiv, 
pp. 5-6, no. 2385; (e) agrees with B. 

Appendix 255 

p" Jude, 1 Pet., 2 Pet.; (b) m/iv; (c) Cologny/Geneve, Bibliotheque 
Bodmer; (d) M. Testuz, Papyrus Bodmer VH-IX (Bibliotheca 
Bodermeriana, 1959); (e) mixed text. 

p" ( a ) Matt. xxv. 43; xxvi. 2-3; (6) — ; (c) Cologny/Geneve, 
Bibliotheque Bodmer; (d) not yet published. 

p 74 (a) Acts i. 2-5, 7-1 1, 13-15, 18-19, 22-25; u - 2_ 4> 6— iii. 26; iv. 
2-6, 8-27, 29-xxvii. 25, 27-xxviii. 31 ; Jas. i. 1-6, 8-19, 21-23, 25, 
27-ii. 15, 18-22, 25-iii. 1, 5-6, 10-12, 14, 17-iv. 8, 1 1— 14 ; v. 1- 
3, 7-9, 12-14, 19-20; 1 Pet. i. 1-2, 7-8, 12-13, 19-20, 25; ii. 7, 
11-12, 18, 24; iii. 4-5; 2 Pet. ii. 21; iii. 4, 11, 16; 1 John i. 1, 6; 
ii. 1-2, 7, 13-14, 18-19, 25-26; iii. 1-2, 8, 14, 19-20; iv. 1, 6-7, 
12, 16-17; v. 3-4, 10, 17; 2 John, vss. 1, 6-7, 12-13; 3 John, 
vss. 6, 12; Jude, vss. 3, 7, 12, 18, 24-25; (b) vn; (c) Cologny/ 
Geneve, Bibliotheque Bodmer; (d) Rodolphe Kasser, Papyrus 
Bodmer XVII (Bibliotheca Bodmeriana, 1 96 1 ) ; (e) agrees fre- 
quently with A. 

p" (a) Luke iii. 18-22, 33-iv. 2, 34-v. 10, 37-vi. 4, 10-vii. 32, 
35-43, 46-xviii. 18; xxii. 4-xxiv. 53; John i. i-xiii. 10; xiv. 8- 
xv. 8 (with lacunae) ; (b) beginning of in; (c) Cologny/Geneve, 
Bibliotheque Bodmer ; {d ) Victor Martin and Rodolphe Kasser, 
Papyrus Bodmer X1V-XV (Bibliotheca Bodmeriana, 1961) ; 
[e) agrees with B. 

p 76 (a) John iv. 9, 12; (b) vi; (c) Vienna, Osterreichische National- 
bibliothek, P. Gr. Vind. 36102; (d) Herbert Hunger, Biblos: 
Osterreichische ^eitschrift fur Buch- und Bibliothekswesen, viii 
('959)»PP- 7-12. 
In addition to the 76 papyri described above, according to a pre- 
liminary announcement 1 official numbers have been assigned to five 
more Greek papyri of the New Testament : p 77 and p 78 are included 
in vol. xxxiii of the Oxyrhynchus Papyri, p 79 is in the National 
Museum at Berlin, 2 and p 8 ° and p 81 (tentative numbers) are among 
the documents found at Khirbet Mird. Still to be assigned an official 
number is a fourth-century fragment of 1 Pet. ii. 20-iii. 12, published 
as no. 2 in the series Papyrologica Castroctaviana, edited by Jose O'Cal- 
laghan. 3 

1 Kurt Aland, Studien zur Vberlieferung des Neuen Testaments und seines Textes 
(Berlin, 1967), p. 91, Anm. 2. 

2 This is presumably the leaf, acquired in 1877-91 by the Library (No. 6774), 
which has now been published by Kurt Treu ('Neue neutestamentliche Fragmente 
der Berliner Papyrussammlung', Archivfiir Papyrusforschung, xviii (1966), pp. 37 f., 
with a Plate. 

3 Serbio Daris, Un nuovo frammento della Prima Lettera di Pielro (Barcelona, 1967), 
with a Plate. 

256 Appendix 

Scraps of New Testament papyri, apparently regarded as too 
insignificant to merit the assignment of official numbers, include the 
following: a sixth-century fragment containing 2 Cor. x. 4 and 1 
Thess. v. 8, 1 a third-century leaf from a miniature codex containing 
some words of the Lord's Prayer (Matt. vi. 10 f.), 2 and a fragment 
preserving a verse of the Gospel of John (iii. 34) .- 3 

The following lists of numerals indicate the distribution of the 
papyri by book in the New Testament : 

Matthew: 1, 19, 21, 25, 35, 37, 44, 45, 53, 62, 64, 67, 70, 71, 73. 

Mark: 45. 

Luke: 3, 4, 7, 42, 45, 69, 75. 

John: 2, 5, 6, 22, 28, 36, 39, 44, 45, 52, 55, 59, 60, 63, 66, 75, 76. 

Acts: 8, 29, 33, 38, 41, 45, 48, 50, 53, 56, 57, 58, 74. 

Romans: 10, 26, 27, 31, 40, 46, 61. 

1 Corinthians: 1 1, 14, 15, 34, 46, 61, 68. 

2 Corinthians: 34, 46. 
Galatians: 46, 51. 
Ephesians: 46, 49. 
Philippians: 16, 46, 61. 
Colossians: 46, 61. 

1 Thessalonians : 30, 46, 61, 65. 

2 Thessalonians: 30. 
Titus: 32, 61. 
Philemon: 61. 
Hebrews: 12, 13, 17, 46. 
James: 20, 23, 54, 74. 

1 Peter: 72, 74. 

2 Peter: 72, 74. 

1 John: 9, 74. 

2 John: 74. 

3 John: 74. 
Jude: 72, 74. 
Revelation: 18, 24, 43, 47. 

1 Herbert Hunger, 'Zwei unbekannte neutestamentliche Papyrusfragmente der 
Osterreichisches Nationalbibliothek', Bibbs, viii (1959), pp. 7-12. It may be 
mentioned that the other fragment referred to in the title of the article has been 
assigned the number p 76 . 

2 Edited by J. W. B. Barns, in Barns and H. Zilliacus, The Antinoopolis Papyri, 
Part 11 (London, i960), pp. 6-7; cf. Ernst Bammel, 'Ein neuer Vater-Unser-Text', 
Zeitschrift fur die neutestamentliche Wissenschaft, lii (1961), pp. 280-1, and 'A New 
Text of the Lord's Prayer', Expository Times, lxxiii (1961-2), p. 54. 

3 R. Roca-Puig, 'Un papir grec de l'Evangeli de S. Joan a Barcelona (P. Bare. 
n° 83)', Analecta sacra Tarraconensia, xxxvii (1964), pp. 353-5. 


The following is a chronological list of introductions to New 
Testament textual criticism. In addition to these volumes many 
dictionaries of the Bible, introductions to the New Testament, and 
histories of the English Bible contain brief surveys of the subject. For 
bibliography on special points of New Testament textual criticism, 
reference may be made to monographs mentioned in the appropriate 
sections above, as well as to two other publications by the present 
writer, namely Annotated Bibliography of the Textual Criticism of the New 
Testament, igi^-iggg (Copenhagen, 1955), and Chapters in the History 
of New Testament Textual Criticism (Leiden and Grand Rapids, 1963). 

J. I. Doedes, Verhandeling over de tekstkritiek des Nieuwen Verbonds 
(Teyler's godgeleerd Genootschap, deel xxxiv), Haarlem, 1844, 
481 pp. 

J. Scott Porter, Principles of Textual Criticism with their Application to 
Old and New Testaments . . ., London, 1848, 515 pp. 

Thomas Sheldon Green, A Course of Developed Criticism on Passages of 
the New Testament Materially Affected by Various Readings, London, 
1856, 192 pp. 

Samuel Prideaux Tregelles, Introduction to the Textual Criticism of 
the New Testament; being vol. iv of Thomas Hartwell Home's An 
Introduction to the Critical Study and Knowledge of the Holy Scriptures, 
10th ed., London, 1856; 13th ed., 1872, 402 pp. 

F. H. A. Scrivener, A Plain Introduction to the Criticism of the New 
Testament, London, 1861, 490 pp.; 2nd ed., Cambridge, 1874, 
607 pp.; 3rd ed., 1883; 712 pp.; 4th ed., 2 vols., edited by 
Edward Miller, London, 1894, 418, 428 pp. 

C. E. Hammond, Outlines of Textual Criticism applied to the New Testa- 
ment, Oxford, 1872, 138 pp.; 6th ed., revised, 1902, 179 pp. 

William Milligan and Alexander Roberts, The Words of the New 
Testament, as Altered by Transmission and Ascertained by Modern 
Criticism, Edinburgh, 1873, 262 pp. 

Frederic Gardiner, Principles of Textual Criticism; with a Graphic 
Table of Uncials, reprinted from Bibliotheca Sacra, xxxii (1875), 
pp. 209-65. 

F. H. Scrivener, Six Lectures on the Text of the New Testament and the 
Ancient Manuscripts which contain it, chiefly addressed to those who do 
not read Greek, Cambridge and London, 1875, 216 pp. 

258 Bibliography 

Thomas Rawson Birks, Essay on the Right Estimation of Manuscript 

Evidence in the Text of the New Testament, London, 1878, 128 pp. 
Brooke Foss Westgott and Fenton John Anthony Hort, The New 

Testament in the Original Greek, [ii], Introduction [and] Appendix, 

Cambridge and London, 1881, 324, 188 pp.; 2nd ed., 1896, 

330, 180 pp. 
J. P. P. Martin, Introduction d la critique lextuelle du Nouveau Testament: 

Partie the'orique, Paris, c. 1883, 712 pp.; Partie pratique, 5 vols., 

Paris, c. 1883-6, 327, 554, 512, 549, 248 pp. 
Philip Schaff, A Companion to the Greek Testament and the English 

Version, New York, 1883, 616 pp. ; 3rd ed., revised, 1889, 618 pp. 
Fr. Schjott, 'Det ny Testamentes Texthistorie i de tre forste 

Aarhundreder', Teologisk Tidsskrift for den danske Folkekirke, i 

(1884), pp. 343-92; 'De nytestamentlige Uncialhaandskrifter', 

ibid., vi (1889), pp. 432-57, 500-55. 
Edward Miller, A Guide to the Textual Criticism of the New Testament, 

London, 1886, 147 pp. 
Benjamin B. Warfield, An Introduction to the Textual Criticism of the 

New Testament, London, 1886, 225 pp. 
Caspar Rene Gregory, Prolegomena, being vol. iii of Tischendorf 's 

Novum Testamentum Graece, ed. octava critica maior, Leipzig, 

1884-94, H26 pp. 
Arnold Ruegg, Die neutestamentliche Textkritik seit Lachmann; ein 

Versuch zur Orientierung, Zurich, 1892, 97 pp. (with a list of N.T. 

MSS., their editors, and collators). 
Eberhard Nestle, Einfuhrung in das griechische Neue Testament, 

Gottingen, 1897, 129 pp.; 2te Aufl., 1899, 288 pp.; 3te Aufl., 

1909, 298 pp.; Eng. trans, from 2nd German ed. by William 

Eadie, Introduction to the Textual Criticism of the Greek New Testa- 
ment, London, 1901, 351 pp. 
Charles F. Sitterly, Praxis in Manuscripts of the Greek Testament, New 

York, 1898; 4th ed. = part 3 of The Canon, Text and Manuscripts 

■ of the New Testament, New York and Cincinnati, 1914, 126 pp. 

Matheus Lundborg, Nya Testamentets text, dess historia och kritiska 

behandling i allmanna grunddrag, Lund, 1899, 406 pp. 
Marvin R. Vincent, A History of the Textual Criticism of the New 

Testament, New York, 1899, 185 pp. 
Kirsopp Lake, The Text of the New Testament, Oxford, 1900, 104 pp.; 

6th ed., revised by Silva New, 1928, 104 pp. 
Frederic G. Kenyon, Handbook to the Textual Criticism of the New 

Testament, London, 1901, 321 pp.; 2nd ed., 1912, 381 pp. 
Rudolf Knopf, Der Text des Neuen Testaments, Giessen, 1906, 48 pp. 

Bibliography 259 

August Pott, Der Text des Neuen Testaments nach seiner geschichtlichen 

Entwicklung, Leipzig, 1906, 108 pp.; 2te Aufl., 1919, 116 pp. 
Caspar Rene Gregory, Canon and Text of the New Testament, New 

York, 1907, pp. 297-539. 
Textkritik des Neuen Testamentes, 3 vols., Leipzig, 1900-g, 

i486 pp. 
Hermann Freiherr von Soden, Die Schriften des Neuen Testaments in 

ihrer altesten erreichbaren Textgestalt, I. Teil, Untersuchungen, i. Ab- 

teilung, Die Textzeugen, Berlin, 1902, 704 pp.; ii. Abteilung, Die 

Textformen, a. Die Evangelien, 1907, pp. 705-1520; iii. Abteilung, 

Die Textformen, b. Der Apostolos mit Apokalypse, 1910, pp. 152 1- 

Jens Stub Irgens, De trykte greske Nye Testamenters historie tilligemed en 

indledning dertil og et anhang, Kristiania, 1907, 196 pp. 
Ernest J acquier, Le Nouveau Testament dans Peglise chretienne. II. Le 

Texte du Nouveau Testament, Paris, 19 13, 535 pp. 
Alexander Souter, The Text and Canon of the New Testament,~Londor\, 

1913, pp. 1-145; revised by C. S. C. Williams, 1954, pp. 1-133- 
P. G. Groenen, Algemeene inleiding tot de Heilige Schrift. II. Geschiedenis 

van den tekst, Leiden, 191 7, 375 pp. 
Ernst von Dobschutz, Eberhard Nestle's Einfiihrung in das griechische 

Neue Testament, 4te Aufl. vollig umgearbeitet, Gottingen, 1923, 

160 pp. 
Heinrich Joseph Vogels, Handbuch der Textkritik des Neuen Testa- 
ments, Munster/W., 1923, 255 pp.; 2etAufl., Bonn, 1955, 236 pp. 
A. T. Robertson, An Introduction to the Textual Criticism of the New 

Testament, New York, 1925, 300 pp.; 2nd ed., 1928, 300 pp. 
Giuseppe Sacco, La Koine del Nuovo Testamento e la trasmissione del 

sacro testo, Rome, 1928, pp. 151-327. 
Ernst Nachmanson, Nya Testamentet, en oversikt av dess yttre historia, 

Stockholm, 1931, 164 pp. 
Leon Vaganay, Initiation a la critique textuelle neotestamentaire, Paris, 

1934, 188 pp.; Eng. trans., An Introduction to the Textual Criticism 

of the New Testament, London, 1937, 208 pp. 
M.-J. Lagrange, Introduction a i 'etude du Nouveau Testament; deuxteme 

partie, Critique textuelle. II. La Critique rationnelle, Paris, 1935, 

685 pp. 
Eugene Mercier, Le Texte du Nouveau Testament, Lausanne, 1935, 

127 pp. 
Auguste Hollard, Histoire du texte du Nouveau Testament, Paris, 1936, 

80 pp. 

260 Bibliography 

Frederic G. Kenyon, The Text of the Greek Bible: a Students' Handbook, 
London, 1937, 264 pp.; 2nd ed., 1949, 264 pp.; German trans, 
by Hans Bolewski, Der Text der griechischen Bibel: ein Lehrbuch, 
Gottingen, 1952, 166 pp.; 2te Aufl., iiberarbeitet und erganzt 
von A. W. Adams, mit einer Tabelle zur Textgeschichte des 
Neuen Testaments von Ferdinand Hahn, 1961, 200 pp. 

R. Wehner, Nya Testamentets grundtext genom seklerna, Stockholm, 

1943. 34 PP- 
Johannes Sundwall, Nya Testaments urtext, Abo, 1946, 104 pp. 
Ernest Cadman Colwell, What is the Best Mew Testament? Chicago 

1952, 127 pp. 
Paolo Sacchi, Alle origini del Nuovo Testamento: Saggio per la storia 

della tradizione e la critica del testo, Firenze, 1956, 178 pp. 
L. D. Twilley, The Origin and Transmission of the New Testament: a 

Short Introduction, Edinburgh, 1957, pp. 36-63. 
Jean Duplacy, Ou en est la critique textuelle du Nouveau Testament? 

Paris, 1959, 103 pp. 
A. I VANOV,'Tekstual'nye pamyatniki svyashchennykh novozavetnykh 

pisanil', Bogoslovskie trudy, i (1959 [i960]), pp. 53-83. 
Vincent Taylor, The Text of the New Testament: a Short Introduction, 

London and New York, 1961, 113 pp. 
Toshio Hirijnuma, New Testament Textual Criticism (in Japanese), 

Tokyo, 1962, 192 pp. 
J. Harold Greenlee, An Introduction to New Testament Textual 

Criticism, Grand Rapids, 1964, 160 pp. 
Bruce M. Metzger, Der Text des Neuen Testaments; Einfuhrung in die 

neutestamentliche Textkritik, Stuttgart, 1966, 272 pp. (a trans- 
lation of the first edition of the present work) . 
Harold K. Moulton, Papyrus, Parchment and Print; the Story of how 

the New Testament Text has reached us (World Christian Books 

no. 57), London, 1967, 77 pp. 

Additional Notes 


p. 3, n. i. For further bibliography on Greek palaeography see Alphonse 
Dain, 'Paleographie grecque', in UHistoire et ses methodes, ed. by Charles 
Samaran (Paris, 1961), pp. 532-52, and Jean Irigoin, 'Les manuscrits 
grecs, 1931-1960', Lustrum, vii (1962 [1963]), pp. 5-93. 

p. 5, line 10. Concerning parchment, see Karl J. Liithi, Das Pergament, 
seine Geschichte, seine Anwendung (= Bibliothek des Schweizerischen Gutenberg- 
museums, vol. vi ; Bern, 1938), and concerning papyrus see Jos6 O'Callaghan, 
'El papiro en el lenguaje de los Padres latinos . . . III. El papiro, como 
planta, en la Biblia: sus alegorias', Studia papyrologica. Revista Espanola de 
papirologia, i (1962), pp. 105-ig. For a general introduction to the study of 
Greek papyri, see Eric G. Turner's recently published volume entitled 
Greek Papyri, An Introduction (Oxford and Princeton, 1967). 

p. 5, n. 1 . See also W. E. Engelkes, Het Grieksche boek in voor-Alexandrijnschen 
tijd (Diss.; Amsterdam, 1926) ; E. G. Turner, Athenian Books in the Fifth and 
Fourth Centuries B.C. (Inaugural Lecture, University College; London, 1952) ; 
A. F. Norman, 'The Book Trade in Fourth-Century Antioch', Journal of 
Hellenic Studies, lxxx (1960), pp. 122-6; and Tonnes Kleberg, Bokhandel och 
bokforlag i antiken (Stockholm, [1962]). The last mentioned has a good 

p. 6, n. 1. For a different view see Saul Lieberman, 'Jewish and Christian 
Codices', Hellenism in Jewish Palestine (New York, 1950), pp. 203 ff., who 
suggests that 'the first Jewish Christians, such as Matthew and Mark, 
would follow the accepted Jewish practice and put down their VTrofiv-^iiaTa 
in codices' (p. 205), for codices or notebooks were regarded as 'the most 
suitable way of indicating that they were writing the Oral Law for private, 
or unofficial use, and not for publication' (ibid.). Cf. also C. H. Roberts, 
'P. Yale 1 and the Early Christian Book', Essays in Honor of C. Bradford 
Welles (= American Studies in Papyrology , vol. i; New Haven, 1966), pp. 25-29. 

p. 8, n. 2. . . For descriptions of Greek manuscripts copied between A.D. 1200 
and 1 400, see Alexander Turyn, Codices Graeci Vaticani saeculis XIII et XIV 
scripti annorumque notis instructi (Vatican City, 1965). 

p. 12, n. 1. Concerning Latin palimpsest manuscripts, see E. A. Lowe, 
'Codices rescripti ; a List of the Oldest Latin Palimpsests with Stray Observa- 
tions on their Origin', Melanges Eugene Tisserant, vol. v (Vatican City, 1964), 
pp. 67-113. 

826158 c 

262 Additional Notes 

For an example of a multiple palimpsest which, according to A. S. Atiya, 
has five layers (early Kufic, over pre-Kufic Naskh, over Greek uncials 
[a portion of 1 Cor. in lectionary form, dating from the late sixth or early 
seventh century], over Syriac [relatively late], over Syriac ['portions of the 
Gospels of St. Matthew, St. John, and especially St. Mark, which [have] 
a considerable number of significant variants from the Peshitta text') ; 
see his article 'Codex Arabicus (Sinai Arabic MS. No. 514)', in Homage to 
a Bookman; Essays on Manuscripts, Books and Printing Written for Hans P. 
Kraus . . . (Berlin, 1967), pp. 75-84. 

p. 13, n. 3. See also W. B. Sedgwick, 'Reading and Writing in Classical 
Antiquity', Contemporary Review, cxxxv (January-June 1929), pp. 93 f. 

p. 17, n. 2. For a revised and enlarged form of the present writer's paper, 
'When did Scribes Begin to Use Writing Desks?', see his Historical and 
Literary Studies, Pagan, Jewish, and Christian (Leiden and Grand Rapids, 1 968) , 
chap. xii. For further discussion of the furniture found in the so-called 
Scriptorium at Qumran, see K. W. Clark, 'The Posture of the Ancient 
Scribe', The Biblical Archaeologist, xxvi (1963), pp. 63-72. 

p. 23, n. 1. Cf. also H. K. McArthur, 'The Earliest Divisions of the Gospels', 
Studia Evangelica, iii, ed. by F. L. Cross (= Texte und Untersuchungen, lxxxviii; 
Berlin, 1964), pp. 266-72. (See also the literature mentioned below in the 
addendum to p. 104.) 

p. 24, n. 2. For further investigation see Carl Nordenfalk, 'The Apostolic 
Canon Tables', Gazette des beaux-arts, lxii (1963), pp. 17-34, an< ^ H. K. 
McArthur, 'The Eusebian Sections and Canons', Catholic Biblical Quarterly, 
xxvii (1965), pp. 250-6. 

p. 25, n. I . For an extended discussion of the Eusebian Canon Tables, see 
Eb. Nestle, 'Die Eusebianische Evangelien-Synopse', Neue kirchliche Z e ^~ 
schrift, xix (1908), pp. 40-51, 93-114, and 219-32. 

p. 30, n. 2. Cf. Ewald Jammers, Tafeln zur Neumenschrift, mit einer Einfiihrung 
(Tutzing, 1965). 

p. 31, n. 1. See also H. Greeven, 'Die Textgestalt der Evangelienlektionare', 

Theologische Liter aturzeitung, lxxvi (1951), cols. 513-22, and Allen Wikgren, 
'Chicago Studies in the Greek Lectionary of the New Testament', in 
Biblical and Patristic Studies in Memory of Robert Pierce Casey, ed. by J. N. 
Birdsall and R. W. Thomson (Freiburg, 1963), pp. 96-121. 

p. 32, line 5 from bottom. The statistics are given by Kurt Aland in his Studien 
ZW Uberlieferung des Neuen Testaments und seines Textes (Berlin, 1967), p. 207. 

p. 33, n. 1 . The most recently published fascicle in the Chicago series is the 
Princeton dissertation of Ray Harms, The Matthean Weekday Lessons in the 
Greek Gospel Lectionary (Chicago, 1966). For a survey of the work done at the 

Additional Notes 263 

University of Chicago on Greek lectionaries, see Wikgren's article cited 
above in the addendum to p. 31, n. 1. The dissertation of Ronald E. Cocroft, 
A Study of the Pauline Lessons in the Matthean Section of the Greek Lectionary 
(Library of Princeton Theological Seminary, 1967), is scheduled for publi- 
cation in the series Studies and Documents (Salt Lake City, 1968). 

p. 33, n. a. For an enlarged edition of the papyrus fever amulet see the 
present writer's Literary and Historical Studies, Pagan, Jewish, and Christian 
(Leiden and Grand Rapids, 1968), pp. 104-10. 

p. 34, n. 1 . For popular accounts concerning the transmission of the ancient 
classics, see Frederic George Kenyon, 'The Lineage of the Classics', Harper's 
Monthly Magazine, cv (June-November 1902), pp. 335-41, and Ernst von 
Dobschiitz, 'Homer und die Bibel, eine uberlieferungsgeschichtliche 
Vergleichung', Neue Jahrbiicher fur Wissenschaft und Jugendbildung, i (1935), 
PP- 33 lff - 

p. 38, n. 1 . Sir Frederic G. Kenyon re-edited the thirty leaves of the portion 
in the possession of the University of Michigan, along with ten leaves 
which he had previously edited, to which were added forty-six newly 
acquired leaves of the same codex, in Fasciculus iii Supplement, Pauline 
Epistles, of The Chester Beatty Biblical Papyri (London, 1936). 

p. 40, n. a. For further studies of p 66 see J. N. Birdsall, The Bodrner Papyrus 
of the Gospel of John (London, i960); M.-E. Boismard in Revue Biblique, 
lxx (1963), pp. 120-37; G. D. Fee, 'The Corrections of Papyrus Bodrner II 
and Early Textual Transmission', Novum Testamentum, vii (1965), pp. 247- 
57; Miguel Balgue in Studia papyrologica, iv (1965), pp. 76-89; E. C. Colwell, 
'Scribal Habits in Early Papyri; a Study in the Corruption of the Text', 
in The Bible in Modern Scholarship, ed. by J. Philip Hyatt (Nashville, 1965), 
pp. 370-89; and E. F. Rhodes, 'The Corrections of Papyrus Bodrner II', 
New Testament Studies, xiv (1967-8), pp. 271-81. 

p. 41, line 7. For further studies of p 72 see fid. Massaux, 'Le texte de la 
I a Petri du Papyrus Bodmer VIII (p 72 )', Ephemerides theologicae Lovanienses, 
xxxix (1963), pp. 616-71 ; M. A. King, 'Notes on the Bodmer Manuscripts', 
Bibliotheca Sacra, cxxi (1964), pp. 54-57; F. W. Beare, 'The Text of I Peter 
in the Bodmer Papyrus (p 72 )', in Studia Evangelica, iii, ed. by F. L. Cross 
(= Texte und Untersuchungen, lxxxviii; Berlin, 1964), pp. 263-5; J. D. Quinn, 
'Notes on the Text of p 72 ', Catholic Biblical Quarterly, xxvii (1965), pp. 241-9; 
and Sakae Kubo, p 72 and the Codex Vaticanus ( = Studies and Documents, xxvii ; 
Salt Lake City, 1965), who concludes that, 'exclusive of singular variants, 
p 72 has as a whole a text superior to that of B' (p. 152). In Jude its text has 
been described as 'wild' and analogous to the Western or Bezan text of the 
Gospels and Acts; so fid. Massaux, 'Le texte de l'fipitre de Jude du Papyrus 
Bodmer VII', in Scrinium Lovaniense. Melanges historiques Etienne Van Cauwen- 
berg (Louvain, 1961), pp. 108-25. 

p. 41, line 15. According to Philippe-H. Menoud, in Acts p 74 agrees more 

264 Additional Motes 

frequently with N and A than with B, especially as to order of words; it 
supports no truly Western reading ('Papyrus Bodmer XVII', Revue de 
theologie et de philosophic, 3 e ser., xii (1962), pp. 1 12-16). 

p. 41, line 6 from bottom. For further studies of p 75 see C. L. Porter, 'Papyrus 
Bodmer XV (P75) and the Text of Codex Vaticanus', Journal of Biblical 
Literature, lxxxi (1962), pp. 363-76; Kurt Aland, 'Neue neutestamentliche 
Papyri', New Testament Studies, xi (1964-5), pp. 5-21, and xii (1965-6), 
pp. 195-210 (reprinted, with additions, in Aland's Studien zur Uberlieferung 
des Neuen Testaments und seines Textes [Berlin, 1967], pp. 155-72); and 
especially Carlo M. Martini, S.J., // problema della recensionalitd del codice 
B alia luce del papiro Bodmer XIV (Rome, 1966). 

p. 45, n. 2. See also Christian Tindall, Contributions to the Statistical Study of 
the Codex Sinaiticus (Edinburgh, 1961) . 

p. 47, line 8. H. Nordberg finds that codex Alexandrinus agrees with the 
dominant type of text represented in the biblical quotations made by St. 
Athanasius ('The Bible Text of St. Athanasius', Arctos, acta philologica 
Fennica, N.s. iii [1962], pp. 1 19-41). 

p. 47, line 21. In 1965 the New Testament portion of codex Vaticanus was 
photographically reproduced by order of Pope Paul VI and copies were 
presented to the members and observers of the Vatican Council II. The 
title-page reads as follows: ra tepa fii^Aia | Codex Vaticanus graecus | 1 209. | 
Phototypice expressus | iussu | PauliPP VI | Pontificis Maximi | H KAINH 
AIA0HKH. J In Civitate Vaticana | 1965. The several forms of the edition 
contain also a second title-page, as well as an Introduction of 21 pages 
(signed by Mgr. Paul Canart and Carlo M. Martini, S.J.) in English, 
French, German, Italian, or Spanish. 

p. 49, n. 2. In the opinion of Hermann J. Frede, however, codex Bezae 
dates from the fourth century (see his Altlateinische Paulus-Handschriften 
[Freiburg, 1964], p. 18, Anm. 4). 

p. 50, n. 2. The book of Acts according to codex Bezae was translated by 
Fr. A. Bornemann in his Acta Apostolorum ab Sancto Luca conscripta ad Codicis 
Cantabrigiensis . . . (Grossenhain, 1848). 

p. 51, n. 2. Codex Claromontanus is dated by Frede in the fifth century; see 
addendum to p. 49, n. 2, above. 

p. 52, line 4. On codex E, see Russell Champlin, Family E and its Allies in 
Matthew (= Studies and Documents, xxviii; Salt Lake City, 1967), and Jacob 
Geerlings, Family E and its Allies in Mark (— Studies and Documents, xxxi; 
Salt Lake City, 1968). 

p. 59, line 17. On the textual affinities of codex Petropolitanus, see Silva 
Lake, Family Tl and the Codex Alexandrinus; The Text According to Mark ( = 

Additional Notes 265 

Studies and Documents, v; London, 1937) ; Jacob Geerlings, Family IJ in Luke 
(= Studies and Documents, xxii; Salt Lake City, 1962); idem, Family H in 
John (= Studies and Documents, xxiii; Salt Lake City, 1963); and Russell 
Champlin, Family Tl in Matthew (= Studies and Documents, xxiv; Salt Lake 
City, 1964). 

p. 62, n. 1 . See also Jacob Geerlings, The Lectionary Text of Family 13 According 
to Cod Vat Gr 121 y {Gregory 54J) [The Farrer Lectionary] (= Studies and 
Documents, xv; Salt Lake City, 1959); idem, Family 13. The Farrer Group. 
The Text According to Matthew ( = Studies and Documents, xix ; Salt Lake City, 
1964); idem, Family 13 (The Farrer Group). The Text According to Luke ( = 
Studies and Documents, xx; 1961); idem, Family 13 (The Farrer Group). The 
Text According to John. (= Studies and Documents, xxi; 1962). 

p. 62, n. 3. See also M. R. James, The Wanderings and Homes of Manuscripts 
(London, igig), pp. 1 7 f. Concerning the scribe Emmanuel, see H. L. Gray, 
'Greek Visitors to England in 1455-1456', in Anniversary Essays in Mediaeval 
History, by Students of Charles Homer Haskins (Boston and New York, 
1919), pp. 81-116, especially 105 ff. 

p. 67, n. 1. For diminutive Greek and Coptic manuscripts see W. H. Willis 

in Classical, Mediaeval and Renaissance Studies in Honor of Berthold Louis Ullman, 
ed. by Charles Henderson, Jr., i (Rome, 1964), p. 270, n. 1. 

p. 67, n. 2. See also the present writer's articles, 'A Survey of Recent Research 
on the Ancient Versions of the New Testament', in New Testament Studies, 
ii (1955-6), pp. 1-16; 'The Early Versions of the New Testament', in 
Peake's Commentary on the Bible, ed. by M. Black and H. H. Rowley (London, 
1962), pp. 671-5; 'Versions, Ancient', in The Interpreter's Dictionary of the 
Bible, iv (New York, 1962), pp. 749-60; and 'Recent Contributions to the 
Study of the Ancient Versions of the New Testament', in The Bible in Modern 
Scholarship, ed. by J. P. Hyatt (Nashville, 1965), pp. 347-69. 

p. 68, n. 1 . See also Allen Wikgren, 'The Use of the Versions in New Testa- 
ment Textual Criticism' , Journal of Biblical Literature, lxvii (1948), pp. 135-41. 

p. Ti, last line. For a convenient scriptural list indexing the several publica- 
tions containing portions of the Palestinian Syriac version, see Fr. Schulthess, 
Lexicon Syropalaestinum (Berlin, 1903), pp. vii-xvi. For additions to the list 
in the book of Acts, see Charles Perrot, 'Un fragment christo-palestinien 
decouvert a Khirbet Mird (Actes des Apotres, x, 28-29; 32-41)', Revue 
Biblique, lxx (1963), pp. 506-55, especially 544. For Syriac manuscripts in 
general, see James T. Clemons, 'A Checklist of Syriac Manuscripts in the 
United States and Canada', Orientalia Christiana "Periodica, xxxii (1966), 
pp. 224-51, 478-522. 

p. 72, line 11. For a rich bibliographical survey see M. Bogaert, 'Bulletin de 
la Bible latine', Revue Benedictine, lxxiv-lxxv (1964-5), pp. [i]-[72]. 

266 Additional Notes 

p. 75, line 5 from bottom. See also Hermann J. Frede, Altlateinische Paulus- 
Handschriften (Freiburg, 1964), and Walter Thiele, Die lateinischen Texte 
des I. Petrusbriefes (Freiburg, 1965). 

p. 76, line 26. According to Westcott and Hort {The New Testament in the 
Original Greek; [ii], Introduction [and} Appendix [Cambridge, 1881], p. 152), 
'the text of A in several books agrees with the Latin Vulgate in so many 
peculiar readings devoid of Old Latin attestation as to leave little doubt that 
a Greek MS. largely employed by Jerome in his revision of the Latin version 
must have had to a great extent a common original with A'. Wordsworth 
and White concluded {Nouum Testamentum Latine, i [Oxford, 1889- 1898], 
pp. 655-72) that Jerome used a type of text represented today by B, N, 
and L. According to von Soden (Die Schriften des Neuen Testaments, I. iii 
[Berlin, 1910], pp. 1544-72) Jerome followed a type of Greek text which 
was similar to the archetype of the three great recensions, /, H, and K. 
H. J. Vogels rejected the conclusions of all his predecessors and held that 
Jerome utilized what modern scholars call the Koine type of Greek text 
(Vulgatastudien. Die Evangelien der Vulgata untersucht auf ihre lateinische und 
griechische Vorlage [Munster i. W., 1928], pp. 55-80). In a review of Vogel's 
monograph, F. C. Burkitt criticized the author for exaggerating the 
influence of the Koine type of text upon the Vulgate, and suggested that it 
is 'more likely that he [Jerome] had at least two MSS., one of which was 
mainly H\esychian\ and the other (or others) mainly K\oine], and that in 
important cases ... he made an eclectic choice between them' (Journal of 
Theological Studies, xxx [1929], p. 412). Lagrange also believed that Jerome 
had availed himself of more than one type of Greek text, but thought that, 
besides codices resembling A and B, he was influenced by the type of text 
represented in N even more than by that in A (Critique textuelle; ii, La Critique 
rationnelle [Paris, 1935], pp. 287 ff.). 

With regard to the Acts of the Apostles, Wordsworth and White (op. cit., 
ii [Oxford, 1905], pp. x-xiii) collected a series of readings which they 
interpreted as showing that Jerome's Greek text differed somewhat from 
any that is known today. After a close scrutiny of these readings, J. H. Ropes 
rejected their conclusion, holding that the Vulgate text of Acts is sub- 
stantially the translation of the Old Uncial text of the general type of B, 
N, A, C, 81, and that of these five manuscripts, it agrees most often with 
A, but also preserves a certain number of Western readings derived from 
the Old Latin (The Text of Acts [London, 1926], p. cxxvi; cf. also Ropes and 
W. H. P. Hatch, 'The Vulgate, Peshitto, Sahidic, and Bohairic Versions 
of Acts and the Greek MSS.', Harvard Theological Review, xxi [1928], pp. 69- 
95, especially 73 ff.). 

As concerns the textual complexion of the Pauline Epistles in the Latin 
Vulgate, according to Lagrange (op. cit., pp. 501 and 509), Jerome seems 
to have reacted against the predominance of the Western type of reading. 

For a survey of research on the Vulgate text of the Catholic Epistles 
since the publication of Harnack's monograph, Z UT Revision der Prinzipien 
der neutestamentlichen Textkritik; die Bedeutung der Vulgata fur den Text der 
katholischen Briefe (Leipzig, 1916), reference may be made to the present 
writer's contribution to New Testament Manuscript Studies, ed. by M. M. 

Additional Notes 267 

Parvis and A. P. Wikgren (Chicago, 1950), pp. 59 f.; cf. also Walter Thiele's 
Wortschatzuntersuchungen zu den lateinischen Texten der Johannesbriefe (Freiburg, 

As for the text of the Apocalypse, H. J. Vogels has shown the resemblance 
of the Latin Vulgate to codex Sinaiticus (Untersuchungen zur Geschichte des 
lateinischen Apokalypseubersetzung [Dusseldorf, 1920], pp. igff.), although, 
strangely enough, he refuses to admit that Jerome based his work on this 
type of text, holding instead that the ancestors of Sinaiticus were influenced 
by the Old Latin (for a refutation of the latter view, see Lagrange, op. cit., 
pp. 609 ff., who argues that not only was the Old Latin base of the Vulgate 
related to that of the Old Uncials, but also that as far as Jerome revised the 
text of the Apocalypse — the amount of revision has been disputed — it was 
still further in the direction of the Old Uncials). The Vulgate therefore 
possesses no little importance in the textual criticism of the book of Reve- 

p. 77, line 20. In the Pauline Epistles the text of the Book of Armagh is 
predominantly Old Latin, with only a small number of Vulgate readings. 

p. 78, line 30. On the Sixtine edition of the Vulgate Bible see Paul M. 
Baumgarten, Die Vulgata Sixtina von i^go und ihre Einfuhrungsbulle, Akten- 
stiicke und Untersuchungen (Miinster/W., 191 1). 

p. 81, line ^ from bottom. An important Coptic witness (copied at the end of 
the fourth or the beginning of the fifth century) to the Western text of Acts 
is described by the late T. C. Petersen in a preliminary article entitled, 
'An Early Manuscript of Acts; an Unrevised Version of the Ancient so- 
called Western Text', Catholic Biblical Quarterly, xxvi (1964), pp. 225—41; 
see also E. J. Epp, 'Coptic Manuscript G67 and the Role of Codex Bezae 
as a Western Witness in Acts', Journal of Biblical Literature, lxxxv (1966), 
pp. 197-212. 

On Greek and Coptic bilingual manuscripts of the New Testament see 
Kurt Treu, 'Griechisch-koptische Bilinguen des Neuen Testaments', in 
Koptologische Studien in den DDR (= Wissenschaftliche geitschrift der Martin- 
Luther-Universitat, Halle-Wittenberg, 1965 [appeared 1966]), pp. 95-124. 

For the Gospel of John in Coptic, see R. Kasser, L'Evangile selon saint Jean 
et les versions coptes (Neuchatel, 1966). 

For the Acts of the Apostles, see Anton Joussen's (unpublished) disserta- 
tion at the University of Bonn entitled Die koptischen Versionen der Apostel- 
geschichte (Kritik und Wertung) (Bonn, 1963 [Vorwort dated 13 June 1959]). 
Joussen's conclusions include the following: '(1) The Greek text pre- 
supposed by the Sahidic and Bohairic (Fayyumic) belongs to that type 
represented chiefly in X and B; (2) Both [Coptic] versions presuppose 
a Greek text that must have contained a mass of Western readings; (3) The 
recognition that the Sahidic and also the Bohairic contain a not incon- 
siderable number of variants which occur elsewhere only in Latin and Syriac 
witnesses shows (as Zahn pointed out, Urausgabe, p. 225) that these transla- 
tions have carried over a not insignificant influence from the same type of 
Greek text which the Coptic versions have followed' (pp. 2i6f.). 

268 Additional Notes 

For i Peter, William H. Willis has announced the preparation of a 
critical edition of that book in Sahidic (see the Yearbook of the American 
Philosophical Society for 1963 [Philadelphia, 1964], pp. 627 f); cf. also 
Willis's article, 'An Unrecognized Fragment of First Peter in Coptic', in 
Classical, Mediaeval and Renaissance Studies in Honor of Berthold Louis Ullman, 
ed. by Charles Henderson, Jr., i (Rome, 1964), pp. 265-71. 

p. 83, line 6 from bottom. The ancestry of the Old Georgian version has been 
disputed (for a survey of earlier opinions see the present writer's discussion 
in New Testament Manuscript Studies, ed. by M. M. Parvis and A. P. Wikgren 
[Chicago, 1950], pp. 41 ff.). According to recent scholars, it was translated 
from an Armenian version that preserved certain Syriacisms; see Arthur 
Voobus, Early Versions of the New Testament (Stockholm, 1954), pp. 187-96, 
and Joseph Molitor, 'Die Bedeutung der altgeorgischen Bibel fur die 
neutestamentliche Textkritik', Biblische Zeitschrift, n.F. iv (1960), pp. 39-53; 
'Zur armenischen Vorlage der altgeorgischen Version des 1. Johannes- 
briefes', Handes Amsorya, lxxv (1961), cols. 415-29; 'Die syrische Grundlage 
der altgeorgischen Evangelieniibersetzung nach Aussage ihrer Harmonis- 
men', Bedi Kartlisa (Revue de Kartvelologie) , xiii-xiv, No. 41-42 (1962), pp. 98- 
105; 'Die Eigennamen in der Johannes- Apokalypse des Euthymius', ibid., 
xvii-xviii, No. 45-46 (1964), pp. 127-31; 'Syrische Lesarten im altgeorgi- 
schen Tetraevangelium', ibid., xix-xx, No. 48-49 (1965), pp. 1 12-18; 
and 'Neuere Ergebnisse zur Textgeschichte des georgischen Neuen Testa- 
mentes', ibid., xxi-xxii, No. 50-51 (1966), pp. n 1-20. 

p. 84, line 10. Other monographs and editions include Ilia Abuladze, The 
Acts of the Apostles according to the Old Manuscripts (in Georgian; Tiflis, 1950) ; 
K'et'evan Lort'k'ip'anidze, The Georgian Versions of the Catholic Epistles 
according to Manuscripts of the 10th to 14th Centuries (in Georgian; Tiflis, 1956), 
translated into Latin by Joseph Molitor in Oriens Christianus, xlix (1965), 
pp. 1— 1 7 ; 1 (1966), pp. 37-45; and Ilia Imnaisvili, The Apocalypse of John 
and its Commentary [i.e., the Commentary of Andrew of Caesarea] (in 
Georgian; Tiflis, 1961), translated into Latin by Joseph Molitor in Oriens 
Christianus, 1 (1966), pp. 1-12 (continued). For the Gospels see J. Molitor, 
Synopsis latina Evangeliorum Ibericorum antiquissimorum secundum Matthaeum, 
Marcum, Lucam, desumpta e codicibus Adysh, Opiza, Tbeth necnon e frag- 
mentis biblicis et patristicis quae dicuntur Chanmeti et Haemeti (Louvain, 

p. 85, n. 2. A preliminary analysis of the textual affinities of the Nubian 
version (so far as it has been preserved) was made by the present writer in 
a paper entitled, 'The Christianization of Nubia and the Old Nubian 
Version of the New Testament', in Studia Patristica, vii, ed. by F. L. Cross 
(= Texte und Untersuchungen, xcix; Berlin, 1965), pp. 531-42, and reprinted 
in his Historical and Literary Studies, Pagan, Jewish, and Christian (Leiden and 
Grand Rapids, 1968), pp. 112-22. 

p. 86, n. 1. On the Anglo-Saxon version, see M. C. Morrell, A Manual of Old 
English Biblical Materials (Knoxville, 1965). 

Additional Notes 269 

p. 90, n. 1 . The Dura fragment of the Diatessaron was re-edited, with a few 
minor alterations, by C. Bradford Welles, et al., in The Parchments and 
Papyri (= The Excavations at Dura-Europos . . ., Final Report, vol. v, part i; 
New Haven, 1959), pp. 73-74- 

p. 91, n. 2. Cf. also Louis Leloir, 'Divergences entre l'original syriaque et la 
version armenienne du commentaire d'Ephrem sur le Diatessaron', in 
Melanges Eugene Tisserant, ii (Citta del Vaticano, 1964), pp. 303-31. 

p. 92, n. 1 . Cf. also the present writer's survey of Tatianic studies in the 
chapter entitled, 'Recent Contributions to the Study of the Ancient Versions 
of the New Testament', in The Bible in Modern Scholarship, ed. by J. P. Hyatt 
(Nashville, 1965), pp. 352 ff. 

p. 95, n. 2. See also R. P. Breaden, 'A Colophon of Interest in the Early 
Printing of Greek', Bulletin of the New York Public Library, 1 (1946), pp. 471-5. 

p. 99, line 9. Most of the manuscripts that Erasmus used in the preparation 
of his editions of the New Testament came from the collection of manu- 
scripts that had been bequeathed in 1443 to the Dominican monastery at 
Basle by John Stojkovic of Ragusa, one of the cardinals created by the 
Anti-Pope, Felix V; cf. Aloysius Krchriak, De vita et operibus loannis de 
Ragusia (= Lateranum, n.s. xxvi, 2-3; Rome, i960), and R. W. Hunt, 
'Greek Manuscripts in the Bodleian Library from the Collection of John 
Stojkovic of Ragusa', in Studia Patristica, vol. vii, part 1, ed. by F. L. Cross 
(= Texte und Untersuchungen, xcii; Berlin, 1966), pp. 75—82. See also Bo 
Reicke, 'Erasmus und die neutestamentliche Textgeschichte', Theologische 
Zeitschrift, xxii (1966), pp. 254-65, who discusses the several Greek manu- 
scripts, now in the University Library at Basle, used by Erasmus for his 
first edition of the Greek New Testament. 

p. 100, n. 2. Bornkamm's view is upheld by Heinz Bluhm in his study, 
Martin Luther, Creative Translator (St. Louis, 1965), p. 75. 

On the question of Calvin's interest in textual criticism of the New 
Testament, see T. H. L. Parker, 'The Sources of the Text of Calvin's New 
Testament', Z e ^ sc hrift fur Kirchengeschichte, 4te Folge, x (= lxxiii) (1962), 
pp. 277-98. In a Postscript Parker states, 'I have since investigated the 
Greek text behind Calvin's Latin. It is clear that he did not confine himself 
to any one printed edition of the Greek New Testament. I am also more 
ready to believe that he made use of Greek manuscripts at first hand' 
(p. 298). 

p. 102, n. 2. The Holy Office declared subsequently (2 June 1927) that its 
decree was not intended to hinder Catholic scholars from thoroughly 
investigating the matter and from espousing an opinion contrary to the 
authenticity of the passage, provided that they profess themselves ready to 
stand by the judgement of the Church ; see H. Denzinger and K. Rahner, 
Enchiridion symbolorum, 28th ed. (Freiburg i. Br., 1952), no. 2198, and 
Enchiridion biblicum, 3rd ed. (Naples and Rome, 1956), no. 136. 

270 Additional Notes 

p. 102, n. 3. From the standpoint of the Greek Orthodox Church, see the 
defence of the genuineness of the passage by Panagotes Ch. Demetropoulos, 
'H yvyjtjLoryjS tov ^uypiov I 'Iwdv. 5.7)8 — 80 nepl tcuv ev rq> oiipavu) 
fiaprvpcuv, in Actes du XII" Congres international d'&udes byzantines, Ochride, 
10-16 septembre 1961, torn, ii (Belgrade, 1964), pp. 429-38. 

In connexion with the history of the discussion of the passage it deserves 
to be mentioned that Sir Isaac Newton wrote at length concerning the 
evidence supporting the passage (which he regarded as spurious) in a letter 
to a friend dated 14 Nov. l6go; see The Correspondence of Isaac Newton, 
ed. by H. W. Turnbull, iii (Cambridge, ig6i), pp. 83-iog. 

p. 104, line 8 from bottom. For wide-ranging information concerning verse- 
division in the New Testament, see William Wright's article on 'Verse' in 
the first edition (the article is shortened in later editions) of John Kitto's 
Cyclopedia of Biblical Literature, ii (London, 1845; New York, 1846), pp. 905- 
14, supplemented by Wright's articles in The Christian Remembrancer, N.s. iii 
(January-June 1842), pp. 455-6g, 672-90. For a list of differences in verse- 
division among about fifty editions of the Greek New Testament, see Ezra 
Abbot's excursus in Caspar Rene Gregory's Prolegomena, being vol. iii of 
Tischendorf's Novum Testamentum Graece, 8th ed., part 1 (Leipzig, 1884), 
pp. 167-82, reprinted in English in Abbot's posthumously published volume 
entitled, The Authorship of the Fourth Gospel and Other Critical Essays (Boston, 
1888), pp. 464-77. 

p. 121, n. 1. Griesbach was also criticized by Frederick Nolan in his two 
volumes entitled, An Inquiry into the Integrity of the Greek Vulgate, or Received 
Text, of the New Testament ; in which the Greek Manuscripts are newly Classed, 
the Integrity of the Authorised [English] Text Vindicated, and the Various Readings 
Traced to their Origin (London, 1815), and Supplement to an Inquiry . . . (London, 

p. 1 26, n. 1 . For a more recent assessment of Lachmann and his work, see 
Sebastiano Timpanaro, La genesi del metodo del Lachmann (Florence, 1963). 

p. 136, line 6. Essentially the same point of view is expressed in the defence 
of the Byzantine text-type of the New Testament by A. Ivanov of the 
Moscow Theological Academy in a series of articles published in %hurnal 
Moskovskoi Patriarchii (1954-56) and summarized by Robert P. Casey in 
'A Russian Orthodox View of New Testament Textual Criticism', Theology, 
1* (1957), PP- 50-54- 

p. 139, n. 1. For more extensive statistics concerning a comparison of 
several modern critical editions of the Greek New Testament, see Kurt 
Aland, Studien zur Uberlieferung des Neuen Testaments und seines Textes (Berlin, 
I9 6 7), PP- 59-8o. 

p. 154, n. 3. See also M.-J. Lagrange, 'La critique textuelle avant le concile 
de Trente', Revue Thomiste, xxxix. 2 (1934-5), PP- 4 00- 9 (— Lagrange's 
Critique textuelle ; ji, La Critique rationnelle [Paris, 1935], pp. 294-301). 

Additional Notes 271 

p. 156, n. 1. For other treatises on textual criticism see V. Coulton, Essai sur 
la me'thode de la critique conjecturale appliquee au texte d'Aristophane (Paris, 1933) *, 
Gilbert J. Garraghan, S. J., A Guide to Historical Method, ed. by Jean Delan- 
glez, S.J. (New York, 1946), pp. 215-31; Studien zur Textgeschichte und Text- 
kritik, ed. Hellfried Dahlmann and Reinhold Merkelbach (Koln and 
Opladen, 1959); Franz Wieacker, Textstufen klassischer Juristen (= Abhand- 
lungen der Akademie der Wissenschaften in Gottingen, philol.-hist. Kl., Dritte 
Folge, Nr. 45; Gottingen, i960) ; G. Thomson, 'Method in Textual Criticism. 
A Tribute to Walter Headlam (1866-1908)', Eirene, Studia Graeca et Latina, 
i (i960), pp. 51-60; Robert Marichal, 'La critique des textes', in L'Histoire 
et des methodes, ed. by Charles Samaran (Paris, 1961), pp. 1 247-1 366; 
Herbert Hunger, Otto Stegmiiller, et al., Geschichte der Textuberlieferung der 
antiken und mittelalterlichen Literatur; Band I, Uberlieferung der antiken Liter atur 
(Zurich, 1961); B. A. von Groningen, Traite d'histoire et de critique des textes 
grecs (= Verhandelingen der koninklijke Nederlandse Akademie van Wetenschappen, 
Afd. Letterkunde, N.R., Deellxx, no. 2; Amsterdam, 1963) ; D. S. Likhachev, 
Tekstologiya ; Kratkii ocherk (Moscow-Leningrad, 1964); and G. Zuntz, 
An Inquiry into the Transmission of the Plays of Euripides (Cambridge, 1965). 

p. 164, n. 1. See also P. S. Coculesco, 'Sur les methodes de critique textuelle 
du type Lachmann-Quentin', Grai si suflet, iv (Bucharest, 1929-30), pp. 97- 
107, and W. P. Shepard, 'Recent Theories of Textual Criticism', Modern 
Philology, xxviii (1930), pp. 129-41. 

p. 169, n. 1. Cf. also Jack Burch, 'The Use of a Computor in New Testa- 
ment Text Criticism', Restoration Quarterly, viii (1965), pp. 119-25, based 
upon Burch's (unpublished) thesis entitled A Critical Study of the Greek Text 
of Second Timothy as Seen in Selected Uncial, Cursive and Lectionary Manuscripts, 
2 vols. (Library of Abilene Christian College, Texas, 1963). 

p. 1 75, n. 2. For an example of the application of such eclecticism, see 
Kilpatrick's chapter entitled, 'An Eclectic Study of the Text of Acts', in 
Biblical and Patristic Studies in Memory of Robert Pierce Casey, ed. by J. N. 
Birdsall and R. W. Thomson (Freiburg, 1963), pp. 64-77. 

p. 177, n. 3. Since 1962 there have appeared the fascicles entitled The 
Pastoral Letters and Hebrews ( 1 963) , and Romans and I and 2 Corinthians ( 1 964) . 
According to H. K. Moulton, the chief criticisms that have been levelled 
against Kilpatrick's text in the Diglot fascicles are: '(1) There has been 
rather too rigid a reliance on word order . . . ; (2) There has been too great 
a dependence on the early versions as against the great uncials . . . ; and 
(3) As a result of these two, and other, considerations the choice of reading 
has been eclectic to the extent that the normal rules of manuscript evidence 
have taken a very secondary place' ('The Present State of New Testament 
Textual Criticism', The Bible Translator, xvi [1965], p. 196). 

p. 177, n. 4. Cf. also Kilpatrick's contributions, 'The Greek New Testament 
Text of Today and the Textus Receptus', in The New Testament in History and 
Contemporary Perspectives; Essays in Memory of G. H. C. Macgregor, ed. by 
H. Anderson and W. Barclay (Oxford, 1965), pp. 189-208; and 'Style and 

272 Additional Notes 

Text in the Greek New Testament', in Studies in the History and Text of the 
New Testament in Honor of Kenneth Willis Clark, ed. by Boyd L. Daniels and 
M. Jack Suggs (= Studies and Documents, xxix; Salt Lake City, 1967), 
PP- i53- 6 °- 

p. 181, n. 2, For other studies of the significance of quantitative and qualita- 
tive analyses of relationships among manuscripts, see J. Fourquet, 'Fautes 
communes ou innovations communes', Romania, lxx (1948-9), pp. 85-95; 
E. F. Hills, 'The Inter-Relationship of the Caesarean Manuscripts', 
Journal of Biblical Literature, lxviii (1949), pp. 141-59; E. C. Colwell and 
E. W. Tune, 'The Quantitative Relationships between MS. Text-Types', 
in Biblical and Patristic Studies in Memory of Robert Pierce Casey, ed. by J. N. 
Birdsall and R. W. Thomson (Freiburg, 1963), pp. 25-32; E. C. Colwell, 
'Variant Readings: Classification and Use', Journal of Biblical Literature, 
lxxxiii (1964), pp. 253-61 ; and E. C. Colwell, 'External Evidence and New 
Testament Textual Criticism', in Studies in the History and Text of the New 
Testament in Honor of Kenneth Willis Clark, ed. by Boyd L. Daniels and 
M. Jack Suggs (= Studies and Documents, xxix; Salt Lake City, 1967), 
pp. 1— 12. For what is described as a rapid method of ascertaining the 
relationships among manuscripts of the later text-types, see Eldon Jay Epp, 
'The Claremont Profile-Method for Grouping New Testament Minuscule 
Manuscripts', ibid., pp. 27-38. 

p. 183, n. 1. For other examples see Fredson Bowers, Bibliography and 
Textual Criticism (Oxford, 1964), pp. 54 f. 

p. 184, n. 1. Cf. also C. Konnecke, Emendationen lu Stellen des Neuen Testaments 
( = Beitrage zur Fbrderung christlicher Theologie, xii, 1 ; Giitersloh, 1 908) . 

p. 186, n. 1. Cf. also H. C. Youtie, The Textual Criticism of Documentary 
Papyri, Prolegomena ( = Bulletin Supplement no. 6, Institute of Classical 
Studies, University of London; 1958), and E. C. Colwell, 'Scribal Habits 
in Early Papyri: A Study in the Corruption of the Text', in The Bible in 
Modern Scholarship, ed. by J. P. Hyatt (Nashville, 1965), pp. 370-89. 

For problems concerning the correct text of Lincoln's Gettysburg Address 
(of which several slightly different copies are extant), see William E. Barton, 
Lincoln at Gettysburg; What He Intended to Say; What He Said; What He Was 
Reported to Have Said; What He Wished He Had Said (Indianapolis, 1930). 

p. 188, n. 3. For a list of errors made by modern editors in deciphering 
Greek papyri, see Youtie, op. cit. (see addendum to p. 186, n. 1 above), and 
his essay, 'The Papyrologist: Artificer of Fact', Greek, Roman, and Byzantine 
Studies, iv (1963), pp. 19-32. 

p. 196, n. 1. The date is that suggested by Tischendorf, 8th ed., in loc. 
Tischendorf is in error in reading to, for tov clearly stands in the gloss ; 
see the plate reproducing the folio of codex Vaticanus in H. J. Vogels, 
Codicum Novi Testamenti specimina (Bonn, 1927). 

Additional Notes 273 

p. 201, n. 1. See also Markos A. Siotis, Ai Soy/jLariKai 7rapaAAayai tov 
Kf.iy.lvov tt)s Kawfjs AiaOriKTjs. A' , TlpoXeyo/jLeva (Athens, i960). 

p. 2 1 2, n. 1 . On the Koine type of text, see the articles mentioned above in 
the addenda to p. 136, line 6, and p. 177, n. 4. 

^.215, n. 2. Add also J. N. Birdsall's study entitled '406, a Neglected Witness 
to the Caesarean Text', Studia Evangelica, ed. by Kurt Aland, et al. (= Texte 
und Untersuchungen, lxxiii; Berlin, 1959), pp. 732-6. 

p. 216, n. 2. On the significance of p 75 in carrying back to the second century 
the type of text represented in codex Vaticanus, see the studies mentioned 
above in the addendum to p. 41, line 6 from bottom. 

p. 224, n. 2. Cf. also Ulrich Becker, Jesus und die Ehebrecherin, Untersuchungen 
Zur Text- und Uberlieferungsgeschichte von Joh. 7. 53-8. II (Berlin, 1963) 
and Kurt Aland in Apophoreta, Festschrift fur Ernst Haenchen (Berlin, 1964), 
pp. 1 1-18 (reprinted in Aland's Studien zur Uberlieferung des Neuen Testaments 
und seines Textes [Berlin, 1967], pp. 39-46). 

p. 229, last line of footnote. On the extraordinary errors of fact (concerning 
palaeographical features of codices X and B) which occur in Burgon's 
appendix on these manuscripts in his book, The Last Twelve Verses of the 
Gospel According to St. Mark Vindicated . . . (pp. 291-4), see Ezra Abbot's 
article, 'On the Comparative Antiquity of the Sinaitic and Vatican Manu- 
scripts of the Greek Bible', Journal of the American Oriental Society, x (1872), 
pp. 189-200. 

p. 230, n. 1. See also J. A. Fitzmyer, S.J., '"Peace upon Earth Among Men 
of His Good Will" (Lk. 2. 14)', Theological Studies, xix (1958), pp. 225-7. 

p. 247, line 10. For further details, see K. Aland, 'Das Neue Testament auf 
Papyrus', Studien zur Uberlieferung des Neuen Testaments und seines Textes 
(Berlin, 1967), pp. 91-136. 


V oS 
3 r~ 

. a 







^ •'. 0- 

■'■* '■.•; .•• vnn •.?.-:- ■ - ..;.;,..■. '■-'■iii&/'- ; 

j i 

Chester Beatty Biblical Papyrus II (p 46 , about a.d. 200), Chester Beany Museum, Dublin. 
Rom. xv. 29-33, xv '- 25—27, 1-3 (see pp. 37-38). Actual size 9 in.,x6 in. 


ere- «j>-JA°y<: ^Knr^)'>aXa;.u=M NiJCT«<? X . 

K)\.C )' Nix- "rAJM e hC*J rij MAf-B^"^' « | 
, , r r /W« e if- w e-jfe j xHrco KXJ f.N ^^j I . r ,, 

oy-'i^E-rfKi e- c ft e-p-'AXH ©7>-J..w"/ r, T / 
.uiw'ni c-j-«-)-c s-r jkrig 'ei eW-fai l f0£?K 

A.T*najA.ycJ K)^J«J^<|^*^*!*& -w ; 

V . ^ BA FAyT-Ui.E-vi oo TTJ ->JT'Y~r I ; 

KiJr-JUU^NfT^f-^Xi.fJrJ-yc-J KAJA__' 'i 

JO»*f»V*y*r i^»t,^WWmwj^r\ -M '■■< 
rrX<iyr) i r-KiaXj^rri >4^Y* : ic..hf^ f ^*-H' V,". 

■I; ;■• ■■' ■ '^ ; f - k ' i ■ '_ i: V • ?''; - «'' i!"' f 'j v 
," ' . ■ '■ ; -i _ ' ' «,. sii'i-', ■*■;.. Ji ; . .. ! M-l l: ii-\ 

f ! 

Bodmer Papyrus XIV (p 75 , about a.d. 175-225), Cologny/Geneva. 

Luke xvi. 9-21 (the name of the Rich Man is given in line 8 from the 

bottom; see p. 42). Actual size io| in. X 5^ in. 


*CllCTHi' I'HCf JAJbr- 

C t M WH ri FTfOCKN 

otuMiAc o aotqm* 

f »»iOKMO#YTOrzef 
4 A Al (flKAl AAAOhA 

-**l *V.T»OICClMur 

F nt-|7ou-friATtu*. 





4 nBJOMfNHCfCfflP 
I* (.-fllTUNMriAAOH- 
" P>i«eN*TOICTNtU 
fcANOlMA©f4TXt- - 
-* -rfffceCTIH* 

e »AA f n eAYTOf*e 


ACAA M)l M Aomiaj 









k ycAi icx Y°n Ml" 


a»l«fYO NTUW4 fX*J 

. a) KKiucOYNAfie 


K AJ OT*f 1 * 'H'#1 1* 



fNeKH OY>4 t -I M-H 
1 1 C'lTOOK Al tfl AK7 
C€ N+OA 1 KTYXt N 

ei c-TM n rn hH*e*K" 
f xe yu>M me rx**" 

6 KATUN r I €NTMK " 
TATT I <■ U r+K A ITO-^ 

Tin NON Til > noy 

iCTYON-Aere i ayt* 


Ticcrct AonrecoT 

Al Alii CI N AYYW* 
TOQ+ Af ro NOM« 
Tf 1 TO N f-i> Af-*«jtW 

3M cere ft nicer: 
Pi e Kf u> nO*bo yn 
i i f icrwyM-Ac r*» 

fuJCIMCD - *"" 

M ei 1 A€0 N TOY"P" 
«« AePKlAYT^^NAf 

& Aim-ex* r^** 

"*• AM OY* nAAiMA*^* 

** MeXereiAarruiMw 
• KfrcYCM AACon 
A AtuceAere i xy 


&> nroiM'AMOY'" 
1 reiAy-itu'ioTp-i* 5 " 


IMU)hf KUANN-i' 
lAeiCMetAY' 11 ' 
ha-Vxh lerroco 




AAmjYn n u» ewr 

r»*Aei VlAYTtOfcCVCip. 

AM 1 1 N AM H N At r- 

:<t wwrencN<M> 


AO|ZU>CUr c 'NCt'- 

r<5»< ifXxyyocAic- 


OA M ATXUA 0"i>C- 1 



en km rA'4*« aw " 
on#T roc'R a«>i i ir~ 

MiUiMTJtMONi u> 
t C^NWNTtl'Av iriM- 

fn iTor-i imocAT 

TPffKAl A^t^l AT!— 


e m emrtu i>o yt- 
Mgt} rAertnAYRi> 

■R>CAN AYMiinn. 
A*i>M*-"NI NfiDlif 

XOMXHll If (.)'•■'■' 

T'»>-'<>aoi ocriei-f- 

A A^A^h l »Y* i °' ) ' ' ' 


CK(-:royKi ! i i K'NAf 

I IOON I u:ki TAW* 
Meye'i N('uM:-'f - 
M Aroy YtK-^t " i< h - 

Ma»Hiih:«imm 'i 

ru>Ni n- r 1 i * 'i i 

KAirfATAC |A| t* 


MArrYriA*y n'Y- 
* r i j nV^kki vkak 

iHi-Al IN>A'AkN» r^ 
J-l I lAINAoi tHiy 

"IAI fAJ-..>M<>J\KJ 
KAlJk: ' 

/ *'YAI T« Ai-M 

/■ KXIA, 


Codex Sinaiticus (x, iv cent.), British Museum, London ; last folio of the Gospel according 
to John (see p. 46), John xxi. 1-25. Actual size 14^ in. X 13^ in. 





4% ± ,jfr«*t^4fcufi*B*c*P» 

s a -S 
^ -- S x 

qj ._ ^ CO 

T3 " 'C 

Sd O 

S C- > 


5 > 
'£> o . 
^ ".5 

u C „ 

:- S ' 

> o 

'- c 

'> tf 


(3 (J 

1/5 ^ 

'-5 CO 

1— 1 CO 

o > 


i ; 1 1 -ft 

•»<(A»'l '/, IV All i 

A A ATM.-, ) J 

I Al 



Codex Rossanensis (Z\ \i cent.), in the Archives of the Archiepiscopa] Curia, Rossano, 

Calabria; folio 8 V (Munoz), Christ and Barabbas before Pilate (with the court stenographer 

standing by the table; see pp. 17 and 59). Actual size 1 1 £ in.xgj in. 




Ht-liifilH.OrC-l DGlVAi A oouc<5 '* 


liiAA° oiAiiTC-OTfOyoYCiAM f 
\:ii(>Y ( -' ro Y Arit, \"- A(J>rc- ii M o 
iirrnci'riC4ik,i4p'i'i ,ic . vc-i'oitu) 
n a|) AAyn i«iv» . coi \(; r tx> oi' Fr 


bo\ icai yu jrOGic Toiioi ICO' ' 

Cov * iCAiiiropoiu-YOOiuc* . I*^* 
j*« • - » i l , . , 

A pa(vi'oni<pALAT on , oyn A6f 
OHAiiTionn Ati'i'tiin - iuCtg 

Oy 1 CtACOaiii ah TACU'ilAO 
.VA'/C-IH'rOiiOU HCl'OM'C ACOTk •■-- 
O \*At3 U O T O D Y 1' WHM3I AO *-» « « »™ 
V JV AlGynAOeilii AM 1 1 IMPA'I'HUBA 


~y IU N *j> Gi^GN \0\^lN M' Ol I 

T0\amJ> a,0 \. IV A OH. M Gil OH G 


* •>• •fl^ -ft-- 






Codex Basiliensis (E, viii cent.). University Library, Basle; Mark ii. 9-r4 (with lidos, lec- 
tionary equipment, and Eusebian canon numerals: see pp. 23, 25, 30, and 52). Actual size 

9 in. X 6-j- in. 



<d Y ■ l^M ^ ' "ft™ (^ |» J»>v-np 

-pnii^roTD)! k^rr^-.kafs 
^> » |jJ \«i -TB«^i -T»- ^tuymi 
vfry <feti)» o^ycy drrmiS^ 
ft-ty -pfe- -p»Uf iT+C.U-ijcivi'' 

neTpoveniCT \ 

nfTfOCVlmcToKocjvx'- ' 
, eKX.6KTOlGtllJ>eniAH 

xumc kiVrmopCcnovi 
TO V- r U,"icr ivok VrmA. 


(fV^cr-.if UZ-pocr. ftvTDMr* 
^Lop 6\jtr -ar^e <p<6 dyjtrfoy 

d -^ Vsfx pjs^Uj -revj^rL»n4y 

KV H xj-imiv ^j. 01C1CVV.T0 

-tnXvXvT6v«;'\tftC V.NV 


Qttr4^MielxM-«L» o-^y <iiix« uj- 
df^wlu&y«U AA/^luci ■>» lv«r 

^ l V*J^*^.T0t»V»»^^f5</4*IJ>«iv«» - *rx* , ^ , W UM / 1 ^ ,, ^*» uv ■ T» - I* -»j.«f: -M>ut. 

|i«y-t^»H'n,l>'tl<^w»VT' ^Yaotai^f^YfaM^^^'^/w^jprU^f.^,. T»..•w,-trf»t*(^$■ 
^. | ,• ll^^»^».»^r*■'T•VT•'^1'£ t '/£ li 'lw'■•»^H' ri^-c^.^yf^f** ^v ^>^U,^v«Vi 

Codex Mosquensis (K"P, ix or x cent.), State Historical Museum, Moscow; i Pet. i. r-2 
and 3, with commentary and scholia (see pp. 27 and 54). Actual size 13J Ln.xgl in. 



(JTfm otm dT , c^Q o u o-l «4> 

•"Uflif I'd" o '{cr * a\\x^Sf 

■id** -• c(mU tuu coi fffr 
do^UJyop a-p<x>»o , 6-y * 5\ 
oLSrourn o'~nnLfa$l$ovo' 

tIclp gySj gLrrot/TOu 
ouf), 6"<jUtU otrtXarnjtJV 

ff*6" i o i/dVq o \ t £ \±f em i\/eu 

Cap A*v° PTt»" * g ccp 

•Toi/Toy eLron>Auot(o* ov 

ujfil ^ t lo 0"n> u Leu «reu 


£ OLUTOp TTBl CO p # cunt 

Ae , v4*TooLoq<Totff;» © 

oup Trt AamiCT'eLLLou 

cxt£P "TO u TO y TO U AO \PV 
I? S *' x ,.~ <** 

ia^v&p ejjeo tom 1 p&e 

luirtH wy axrlTou uJUiuw 
*TB<«Tmn)X€V e t 1 ^ 

P°P Al«DO«TpCO«T«p g 

.£#* J "V* ^ Y*mju>u4ou" 
Toviunjjsvaj .coo ou »■ j» 

TO I «• 1 «j^CM o l cr- * t y'cr' 6 

iLAcd\eVo-Cip-cop v#V. 
J s o-^tpctU/yeu^Qjj # 'etf>3^ 

af o y crp <d ff° v CUJ * ro K *•* 

\ g yf*OJUT0 1 Cf 0"TDiX«tJ 

Toer * Toi/aeurtAfeouVj' 

Uico^ari^coaro -•*• at-nro* 

Lp 1 •^Lcroifr o l cLfl yl 6y <cr * 

ou(l4xsLL^>vuaJaiX&cu , 

V 1^1 ^ * '' 

q mU UoneroLpac.' to, 

T^owp TBTtpe^CO"-^/ 
ctu Top a-i/TOiojiiiCLeijpetst 

*#*•:• cvArreAi one- 

KATA M AT *(hA I OH •'.• 
p co •;• j i^D y \ o vb**:,. 

p^Ta^ta-AW^K* 1 "^ au 
1 «TOTBT» 6'tovTourw 1 
9 y tou clp v" p 1 eu 
cr Of vt CTD (M (Tl ^ 
', flw cotaaj* 1 6" u 01 c 
A«n«Djj • IvpjtfToy 


r ptoo cryTcoycuD 4 «uierv<» 

VAT. GR. N"^.. 


Greek Gospel Lcctionary 562 (written, according (o a colophon, in the city of Capua, 
a. d. 991), in the Vatican Library, Rome; John xix. 10-16 and Malt, xxvii. 3-5 (see pp. 3of.). 

Actual size 9J- in. X -]\ in. 


• ' ■ 
v ... \ v/ 

Hoy flunv UtV 11 ^ " tnu^fr { -rrcu-rtr t^ia* 
lij*sfrb\fTBA c cowrou /< <M "too -ttv -rp oa 6 *W 
CT» ovy^ Cfl^uf <f crlMUJ v-A • A <M of ^-fr-fc ^ **• 

" ^ .e. . > • ^ ^ - , • ; i °<^ 

no CO I TTP OO -TV\ OTtJU_) O-OJ-TT) \j 6-J> OJ-» U. 1TB 00 » JT» rfr* } * I 

no p 6-u -6^ cttu olj ^-uS-U/sA/l' >*~£y tt i cr ^i^tt*u 
_, e \jtri y,+. talu( y«i to tuo «j cttu^j ■ nfo ' o -n^U 

to\ji y 6- «F«-p frp oo' -ft-U C-y frff •" Y^ P. 4*^ 
ttd p fr-v o yiX^/ o i cr 4 croi^p o Ufa K. eijc^ y% t 
Jl nj-Tirfr-A "^op -nfcr o-> ~-ur1*riA/( A-a-V ***' •" 'V>o i t>8v«. 

S 1^) U *« S^t^ f/ I cr (j-rrl qT$J3TILp ^. O C|^p o J/O"- 5 
. *— • ■»• fc^ 

'**^ »(r"-*| »' -*^ ICola oo ^ <f Aa <rffv t^w *«J 
TTl <rl OLp •ULn-oo p Iuxa ^< Xiv^> o tuxf <Aa' ay 
^* *r o "H *ifc i cr -*wt^rrt^tctAj o i er eVrrro p C\/^VV^P t^V*"* 

«u p'tT|» m ty ht iuvuct ynr r «i »»m.u(T4W TifTiJf 

Greek Gospel MS. 274 (x cent.), Bibliotheque Nationale, Paris; Mark xvi. 6-20 and the 
intermediate ending of Mark (see pp. 30 and 226). Actual size g| in.x6| in. 




«*B B ■* 
01 ^ '«■ ~. 

^. £.3 H.2L2 

— a, *-. 

I as I 

S j* » SL 


irz /j 

J u £ * u y . £ 
B |,= 2^| S& 

^. .ii •" ■Tfi.^*.\ii( 

£ ■ " ■ * "2 aj. 




^ ^^ " — 

* ?«€■ 


w £9-</t £ 2 fl 

W:S /I £ 

IB £ -4 2. 

•>-4 z s 

"2 -* 1 - > 

»* D £ 

i«- ~" JB 

* £-U- 

E E ^B Z*~t 

1 st-z 4 - - ^ 


5 £ £* >2-S j* 

aL.3 >d! 

Sp** ITS Z.= 


< Q. 

i iflffitl 

' JHiSjfllffl*. 




3 C 

c/3 r 

3 O 


Hi | cfg 


? 8 P 

• .? i? U -1 

> VI -- - - ■ 
»■«'. si? 4i 

=i * 

3 £ 

elsf fl .lit! 

tjr.a. 1,1,413 if 

3 j c t :r~\t%< 




a i 5 £ / fris ■-' $3 « 8 * s 

'sis , t EsSS-j^ 



it & 



■* >3 


■ 1 


5 3 ** 



*.l ^s >& '4 

Ik vj i 

-a f 

j to 

■"■a - 4 f- s 
Ills nh* 


P- 1 



- dob 

•^ X IT) 

a; ^ D- 

— ^ 0) 

S u 5 
bo V £ 


. . -d 

■a Js 




I ' >]« S) 


t 'H'l"oa"*(ilT6»lcq«*>Oui|H 
Alt IUjJL)0:AYU>T1uy4XeH^<f 

■$- .t"-MGYnay^eri6n«v*x«-iiA 

* tc|** i ie * AAYc*»*mu; • ri e n-rxM 
* ;c*»«»i icr^AVKi^H'rknieriiutr ■ 

> ZWj 1<AKC A Yt«.Ul ftl f EK4K* 

M6eAVTM^oovq 6 RaXpV-iu 
i »^oY'»-eenetjf AMner^pA' V 

,Jf6K4oeqefUMTPe- vj-A' 
>»>>>.^n>>- eTKenojroiM' 


oeiM' » KivofiuUuerTiJ'/ 
OFiMi'i'iimt Hiwnr-r.^Hii; 
GrtKnr i«(ic- Nf 'fr uriKOc 
ttt» i;n( ■ ay «»»mw %«ti Kiirtiur 

(a) Chester Beatty Coptic (Sahidic) MS. B (about 
a.d. 600), Chester Beatty Museum, Dublin; John i. 1-10 
(see p. 80). Actual size 4J in. X 4 in. 


'-""HJJ1J.-,J. F 


n?ou-irt>o. itnmft 
•rtim.'i-NnM H-ikwrf^f 

•* lirMiuntr/rbntifrirh 

•|4!Man>i.«niPi;p.«Ms k , 

k *fc£HIIIU01>-|, • ph- !,-«£, 

-'iifiqi-n|t«ttru*iivi iu(' 


(6) Armenian Gospel MS. 229 of the Patriarchal Library of Etchmiadzin (dated a.d. 989), 

now in the State Repository of Manuscripts, Erevan; mention of Ariston the Presbyter, 

col. b, between lines 6 and 7 opposite decorative boss (see p. 227). Actual size of entire 

folio 13$ in. X iof in. (plate reproduces lower three-quarters of folio iio v ) 


rlf't' i^*ic 




* frn'jfflil . .. ". M-v2-' ^ - -' I '" -J- 

* *"«n*#^-A«Mt«fl«lOlUXjf« / ;.r<^/»iujO-j/ V fi\« ll/ ^ 

p o/**trirBootjUK.«4<TCuJt>v«>ri|/-r»o/a«««»t4«B|i 

{itotrturmjof/ioD^s^^uTBurouiJoio-' Ko/nu 
"T»ura^*f if w» i o tui to i «arp o ft |i «jtu o- «i "an> «<r ouu 

k -n» -n; -otvi ■Mfi^i1b > 'ra / 


W'-J,:\ l 

■ A -...--Mm 

Greek Gospel MS. 2 (xii cent.), University Library, Basle; Luke vi. 20-30; one of the 
inferior manuscripts used by Erasmus for his first edition of the Greek New Testament 
(see pp. 98-100), with his corrections and annotations for the printer (e.g. in lines 8 
and 9 an oblique stroke separates the definite article from the following word, and in the 
lower margin is Erasmus's addition of TrpootvxtaOt vrrkp rwv eV^uea^dvTOjv upas, which 
the scribe had accidentally omitted from the text of vs. 28, third line from the bottom). 

Actual size 7$ in.x6 in. 




•a<rTct>^l'^^^ , E<WTOl'(; , a^o?va^lBavoWTt<;.• koh' kg 
e<fc<;'ov/K'EAoKixia<Jaynou T e6Qp'6X6ip'E|j ,, 6niY 
ucixJEi < Trap82kCDK£|J''avTou<;/0'eE6<;'Ei<; , a2i6KHiop 
"l>ovv>'c»iEipna'uH , KaeHKoyTa,"'raEroXHpociju£ 
^uc/OTd<JH°aZiKi'a, p E:oppEia,' , raopHpia,'n}iEO 
' ZoAou , * KOK0H6E1CH . vJ/iev/piffTat; . ' KcrrccAd 
7vou<;. , '»Eo<rrvyY6i<;,'vBpiffTd(;i'\/TrepH<})d(Jovy<;. , a 
>va36pa(;, l! sq)iupETd(;'Kcn:flbp,'yo|JS\>cnij , an:i6«i(;. 
"aavpSTOv<;."aavpeETOvc>°a<yTdpyovc; F 'aan6p 
'eEovy En!ypd|/TE<;.'6Ti/oi/7a , 1oiav>Ta'ropd<;« 
aoixTE<; a d|ioi 'eapdrrov/'EKJip, " ov' judpop 'av/Ta 
'noioOop.* a^iXd'xai ' avpEv2LOKOvaip/7oi<; 
'npdaaovai. Cap.i. 

[WBi6 b a(jaTroAdyHTo<;' Ei'ao 'd^epconE 'nat;/o 
'g/$ l 'Kpi\xo\>.''t\i'a>'Yap'Kp\\i&\i m ZTipo\),"az 
auTd(/*KorraKpipEi(;. iia' yap'avTd ' npdo- 
Cttoo'upjpooy.' oi'2iaju8ii"2iB'6Ti /7o' Kpiuanov 
'OtoO'EtfTi'KaTa'aAHflsic^'OTnovqAa'Ipiav : 
Ta'npda<JoyTa<;. , AoyijH' , 2kEHouTO K a>'dtjepcuns 
/o" Kpipa>y/"Jov<;/7a''"Toiav/Ta°npd<j<JopTa<; p koi 
'nioicoif'auTd ,*oti'<jv/"ek(Pe\>^h /7o'Kpijua/7ov 
Tou"Kai/lH^'a|joXH<;'Kai/lH(;'JuaKpoevJii'a<;' l Ka 
TacppopEu^ayyoaJp 'dTi/"Io'XpHaTdpv7<5\/ *e.E= 
o\>"£i(; <> jiETd(;oiajJ , 06' , dyEi. , KaTd"2*EnHp'OK>iH 
poTHTa'aou'Koa'auETaiidHTOvi ' Kap2u ap ' e Haau 
pi]Si<; ' osauTcb 'opyHp'tp' mispa'bpyH<; ' Kqi "a 
iTOKaAv\{/6co<; ' nai ' AiraiOKpiqiac, / 7ov ' 6eo\>. 
'^ol<?ll.l\>'Ka«'\J■^o^l.o\)H\>'ipYO\J , ayaeov 'A6 
fa[j , Kai ,, '7iuHiJ'Kai , 'a99apciap'E7TijHTo\>ai»'3cD= 
Hp , aWbpioiJ. l! ioi(;'^E > Ef l £pieEia(;," , Kai*aTreieoO 
ci°jusp/7H f aAH«Eia>' 1 03Eieo.ueVoi(; r AE/7H'a2ki- 
Kia, 'euaio^'Kai'opyH ' flAi v{/iq ' Kai ' <mpoXco -- 
pia b em' c rodcra|j'''4'uXHp'a|jepcbnov/ 7ov/V,a^ 
l Kai'"E>iAHiJO(;."2idfa 2iE'Kai''7iJUH'Kai 'EipHpH 
Ht'cjpajTop 'koi'sAAhui ."ov/'yap'effTi'ropo* 
aoonoAH\|/ia ( Gjapd/7a>'eEa> . k d<roi ' yap "apd^ 
iicoq " HjiapTO)/. • apd.u<»<; ,, Kai'>aTTo AovpTai . ' Kai 
• 6001' Ep*po.u.oo'HJUiapTop,' Aid' pdjuov ' Kpiewrop 
Tai. K ovy'Yap.oi°aKpoaTai'/7ovy'|jdiiou'2iiKaioi 
•roapd 7co b 6Eco , ' aAA/01 ' otoihtoci / 7ou ' yjouou 
"Zii;aia>OHaoiJTai."dTa|J''Yap , 'eeiJH 7a".UH r pd-= 
.uop*EXopTa'.9uaEi"7a/7ov'pdjuiov/raoiH. , o\);- 
Tips<;°EpZeiKpupTai,7o'E'pyop(7ov'pdjuiov 'ypa 
jrrop m Ejj/7ai<;" capiiaiQ'auTobp . ' ffvjajaapTu •' 
pov<yHr > ''av/Ta>p "Ihc, 'aupEiiHaEffix;, ' Kai "JiitTa 
j\>"aAXHA«op /Jcop ' Aoyicmcop ' naTHyopovp^ 

Complutensian Polyglot Bible, vol. v. first printed Greek New Testament (1514), Rom. 
i. 27-ii. 15 (see pp. 96-98). Actual size 14 in. xg^ in. 

O fiu p in , femctip(io 'recipiciitcc . ' Et ' licut 
°non'p!ol»3ucrflt'ocu'Nbir£ , tii , 'iionria: 
'tradidirilloo'ticiis'in'rcpiobum ccooo 
'ffflfuj'vi'fariant 'council 'couciiiu^'re 
pletoe"oi°iniqiuiatc:''mauaa: > fo:nicarto 
lo?: , pi-nnbiio l ii6 l olx"dicK(:"'iivfipicntcs: 
"incompofuofl:°fi^affccrion« p abfq} | 'fc 
dctt: q finc q mia. r ^liii''CLi*iiiftina'tfi"cogno 
gni'fut k moi[c:°iio'foluq'ca'faaut: , 'fcd'e[ 
qui^confcntiunt'faciciiribuo. Cap.j. 

la^l 'oio'i|'nidicae.'''gn'quo'ci , "alfcy'ui a*.-.,, 
dicae" tcipfum 'codcnao .' £a<icm ' cnim 
'agio*qiu:' , iui)ic3s. , Scini'?"cni'qinMudi 
cium ' oci ' cfl 'fctfrn 'vcritatcin "111 'coo'q 
"taua'agut.'£iiftimais , 'aut'boc'o' bomo 
n 'qui ,, 'iiidica8'coo°qiii"t3lia agiini r ct» 
"facie'cayquia'tii-cffiigice 'iudiciiini a> 
'octf'Sn'DiLUtiao'bonitarie'cuio ccco 
' 5sno;a8 k quoniam ' bcmsnitas^oa oooa 
"ad'pniam ! 'tCaddiicu; , Scc)m'aut'oun= 
'nWiranTiiL'eic'irc'ct'reutlarionio 000 
'iufli'iudidi'oci: aocroccocccooxocw 
n, qm"'rc<lda'vmciiiqj , 'fcamdu5' , opa , d'. 
■feii'qdc'q'fc^' panctiaboru'opis 'glo= 
na}'-J , 'boiio!^'t°irKo:nipnonc'circnb'''rf * 
li'«erni*foifl'liqfQt > a- l ?tac>£'''rq"n6 
"acil'fdit p vcntan' 1 crcdiini'aut"iiuquitari: 
b m'omiKin' > animam'lx>rntiii0'operann8 
•malum "iiidci "pninumoceoax'occocco 
'omm"opcrarurboiium:'indco coxcooo 
■ pamaVsrcco. "flo* ci'c(Va«q>no'p1b» j* ' ' t 
nap'apud'Pci'i.'Oincikp , cni"'fuic"Ug« t^. < 
"pccc»urri'u:' , finc°Icgi: , pcnbum. 'Etcw *£\ t 
biinf.'"Bon r ciii'aih1i[o:i:e'Ic0io'iuflifuiit n?» • • 
'aputl'ecfi'.'fcd'facto^a'lcgiococcaxo **•" ' 
"uiftlficabiinf •' £u'nii 'ijcrWq'ICijo*'"' 
•babci'n9niralirfr"ca"q'li"5ii"'lui'fai.'iu i: 
'«ufmodi J kgc , 'iio'l>JbC[i"!3''ipfi*i"ibi'fui 

'olkndui'opuc-'U'ijii"' Tcnpnim l'Iit:''! 1 " 
" , iircoJditi'''liiw. r icltimoniu l 'rcdt1<iciUn 
, confaintia q ipfo:um:'ci'inriTca''CWCO 
"fc* inui^aii ' coQiiJtiomun ' jiaifannum